--------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\hanna\Dropbox\PC\Documents\PaperProjects\Paper-Effective resistance\Code\ReplicationMaterial\Output_Append
> ices.log
  log type:  text
 opened on:  17 Oct 2019, 10:14:57

. 
. set more off

. version 13.0

. clear

. 
. //cd " " // Change to the directory where you saved the data and the VarLabels.do
. 
. use "dataForAnalysis_v3.dta", clear
(Written by R.              )

. 
. destring YEAR, replace
YEAR: all characters numeric; replaced as int

. xtset cowcode YEAR
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

. 
. ** Labels
. do "VarLabels.do"

. 
. 
. 
. label var RESTRICT_COUNTdomlag1 "Restrictions"

. label var PTS_Slag1 "Political Terror Scale" 

. label var hrgroupslag1 "Human rights CSOs"

. label var hrnewslag1 "Human rights news"

. label var protest_ClarkRegan_loglag1 "Protest count"

. label var UCDP_armedConflictlag1 "Armed conflict"

. label var PR_freedomHouselag1 "Political rights"

. label var PR_freedomHouselag1_sq "Political rights sq." 

. label var gdp_pc_constantUS2010lag1 "GDP per capita"

. label var gdp_pc_constantUS2010lag1_sq "GDP per capita sq." 

. label var KOFGIlag1 "Globalization"

. label var KOFGIlag1_sq "Globalization sq."

. label var populationlag1 "Population size"

. label var urgentActionslag1 "Urgent Actions (lag 1 yr)"

. label var urgentActions "Urgent Actions"

. label var shamingINGO "INGO shaming"

. label var shamingINGOlag1 "INGO shaming (lag 1 yr)"

. label var fhbest "Political rights best"

. label var fhworst "Political rights worst" 

. label var deathpenalty "Death penalty"

. 
end of do-file

. 
. 
. ***************************************************************************************
. *** Appendix S4: Different operationalization of independent variable "restrictions"
. ***************************************************************************************
. 
. ** Leave out on restriction type at a time
. 
. rename REGISTRATION_PROBLEMS REGISTRATION

. local varlist "SOME_BANNED VISIT_RESTRICT TRAVEL_RESTRICT FUNDING_INT FUNDING_DOM REGISTRATION CENSOR HARASS_AMOUNT_bi ARREST_
> bi SURVEIL_bi KILLING_bi"

. 
. foreach var in `varlist' {
  2.          
.          sort cowcode YEAR
  3.          gen `var'lag1 = `var'[_n-1]
  4.          gen RESTRICT_`var' = RESTRICT_COUNTdomlag1 - `var'lag1
  5.          gen RESTRICTsq_`var' = RESTRICT_`var' * RESTRICT_`var'
  6.          label var RESTRICT_`var' "Restrictions (-- `var' )"
  7.          label var RESTRICTsq_`var' "Restrictions sq. (-- `var' )"
  8.          gen RESTRICT_`var'l2 = RESTRICT_`var'[_n-1]
  9.          gen RESTRICTsq_`var'l2 = RESTRICTsq_`var'[_n-1]
 10.          gen RESTRICT_`var'l3 = RESTRICT_`var'[_n-2]
 11.          gen RESTRICTsq_`var'l3 = RESTRICTsq_`var'[_n-2]
 12. 
.         * Negative binomial 
.         xtset cowcode YEAR
 13.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_`var' RESTRICTsq_`var', vce(cluster cowcode) ;
 14.         #delimit cr
delimiter now cr
. 
.         estat ic
 15.         mat es_ic = r(S)
 16.         mat list es_ic
 17.         local AIC: display %4.1f es_ic[1,5]
 18.         local BIC: display %4.1f es_ic[1,6]
 19.         local LL: display %4.1f es_ic[1,3]
 20.         outreg2 using ".\Tables\Appendix_TableS4_`var'.doc", replace ///
>         keep(RESTRICT_`var' RESTRICTsq_`var') ///
>         ctitle("Model 1:") label eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
 21. 
. 
.         * Negative binomial
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_`var' RESTRICTsq_`var'
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, vce(cluster cowcode);
 22.         #delimit cr
delimiter now cr
. 
.         estat ic
 23.         mat es_ic = r(S)
 24.         local AIC: display %4.1f es_ic[1,5]
 25.         local BIC: display %4.1f es_ic[1,6]
 26.         local LL: display %4.1f es_ic[1,3]
 27.         outreg2 using ".\Tables\Appendix_TableS4_`var'.doc", append ///
>         keep(RESTRICT_`var' RESTRICTsq_`var') ///
>         ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
 28. 
. 
.         * Zero inflated negative binomial 
.         #delimit ;
delimiter now ;
.         zinb urgentActions RESTRICT_`var' RESTRICTsq_`var'
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, 
>         inflate(fhbest fhworst UCDP_armedConflictlag1 deathpenalty urgentActionslag1) 
>         vce(cluster cowcode);
 29.  // vuong ;
>         #delimit cr
delimiter now cr
. 
.         estat ic
 30.         mat es_ic = r(S)
 31.         local AIC: display %4.1f es_ic[1,5]
 32.         local BIC: display %4.1f es_ic[1,6]
 33.         local LL: display %4.1f es_ic[1,3]
 34.         outreg2 using ".\Tables\Appendix_TableS4_`var'.doc", append ///
>         keep(RESTRICT_`var' RESTRICTsq_`var') ///
>         ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
 35. 
. 
.         
.         * GMM
.         xtset, clear
 36.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1
>         ( RESTRICT_`var' RESTRICTsq_`var' = 
>                 RESTRICT_`var'l2 RESTRICTsq_`var'l2
>                 RESTRICT_`var'l3)
>                 , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
 37.         #delimit cr
delimiter now cr
.         
.         estat overid
 38.         mat es_ic = r(J) 
 39.         matrix list es_ic
 40.         local J: display %4.1f es_ic[1,1]
 41.         outreg2 using ".\Tables\Appendix_TableS4_`var'.doc", append ///
>         keep( RESTRICT_`var' RESTRICTsq_`var') ///
>         ctitle("Model 4") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Sargan-Hansen-test, `J')
 42. 
. 
.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq 
>         hrgroupslag1 hrnewslag1 
>         populationlag1  UCDP_armedConflictlag1 
>         (RESTRICT_`var' RESTRICTsq_`var'
>         PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
>         protest_ClarkRegan_loglag1 = 
>                 RESTRICT_`var'l2 RESTRICTsq_`var'l2
>                 RESTRICT_`var'l3
>                 PTS_Slag2 
>                 PR_freedomHouselag2
>                 PR_freedomHouselag2_sq 
>                 protest_ClarkRegan_loglag2 )
>         , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
 43.         #delimit cr
delimiter now cr
. 
.         estat overid 
 44.         mat es_ic = r(J)  
 45.          matrix list es_ic
 46.         local J: display %4.1f es_ic[1,1]
 47.         outreg2 using ".\Tables\Appendix_TableS4_`var'.doc", append ///
>          ctitle("Model 5") label  eqdrop(lnalpha) dec(3) ///
>          keep(RESTRICT_`var' RESTRICTsq_`var') ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Sargan-Hansen-test, `J')
 48. 
. }
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5238.9863  
Iteration 1:   log pseudolikelihood = -5238.9591  
Iteration 2:   log pseudolikelihood = -5238.9591  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2663.0312  
Iteration 1:   log pseudolikelihood = -2644.5648  
Iteration 2:   log pseudolikelihood = -2632.6321  
Iteration 3:   log pseudolikelihood = -2632.5753  
Iteration 4:   log pseudolikelihood = -2632.5753  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      64.66
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2632.5753               Pseudo R2         =     0.0419

                                        (Std. Err. adjusted for 171 clusters in cowcode)
----------------------------------------------------------------------------------------
                       |               Robust
         urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------+----------------------------------------------------------------
  RESTRICT_SOME_BANNED |   .7526224   .1196305     6.29   0.000      .518151    .9870938
RESTRICTsq_SOME_BANNED |  -.0677243   .0142481    -4.75   0.000    -.0956501   -.0397985
                 _cons |  -.1920383   .1699349    -1.13   0.258    -.5251046     .141028
-----------------------+----------------------------------------------------------------
              /lnalpha |   1.249097   .1047892                      1.043714     1.45448
-----------------------+----------------------------------------------------------------
                 alpha |   3.487191   .3654199                      2.839743    4.282254
----------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2632.575       4    5273.151   5294.883
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2632.5753           4   5273.1505   5294.8828
.\Tables\Appendix_TableS4_SOME_BANNED.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12091.305  
Iteration 1:   log pseudolikelihood = -6477.5694  
Iteration 2:   log pseudolikelihood = -4229.9746  
Iteration 3:   log pseudolikelihood =  -3007.716  
Iteration 4:   log pseudolikelihood = -2920.5392  
Iteration 5:   log pseudolikelihood = -2918.8004  
Iteration 6:   log pseudolikelihood = -2918.7994  
Iteration 7:   log pseudolikelihood = -2918.7994  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2010.1066  
Iteration 1:   log pseudolikelihood = -1859.8043  
Iteration 2:   log pseudolikelihood =  -1847.418  
Iteration 3:   log pseudolikelihood = -1847.3223  
Iteration 4:   log pseudolikelihood = -1847.3223  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     275.54
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1847.3223               Pseudo R2         =     0.1364

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_SOME_BANNED |   .4650487   .1089525     4.27   0.000     .2515058    .6785917
      RESTRICTsq_SOME_BANNED |  -.0396528   .0128803    -3.08   0.002    -.0648977    -.014408
                   PTS_Slag1 |   .6031834   .0962124     6.27   0.000     .4146107    .7917562
                hrgroupslag1 |   .0018023   .0026784     0.67   0.501    -.0034474    .0070519
                  hrnewslag1 |   .1965428   .0550768     3.57   0.000     .0885942    .3044913
  protest_ClarkRegan_loglag1 |   .3098174   .0933054     3.32   0.001     .1269421    .4926927
      UCDP_armedConflictlag1 |    .376638   .2466029     1.53   0.127    -.1066948    .8599709
         PR_freedomHouselag1 |   .5853659   .2348093     2.49   0.013     .1251481    1.045584
      PR_freedomHouselag1_sq |  -.0552621   .0285822    -1.93   0.053    -.1112823    .0007581
   gdp_pc_constantUS2010lag1 |   .8029638   .3362054     2.39   0.017     .1440132    1.461914
gdp_pc_constantUS2010lag1_sq |  -.2752101   .1271031    -2.17   0.030    -.5243276   -.0260926
                   KOFGIlag1 |   .1035691   .0444619     2.33   0.020     .0164253    .1907128
                KOFGIlag1_sq |  -.0009219   .0004206    -2.19   0.028    -.0017463   -.0000975
              populationlag1 |   .0712058   .1720085     0.41   0.679    -.2659245    .4083362
                       _cons |  -6.109849   1.279407    -4.78   0.000    -8.617441   -3.602258
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5299482   .1218098                      .2912053    .7686911
-----------------------------+----------------------------------------------------------------
                       alpha |   1.698844   .2069359                      1.338039    2.156941
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1847.322      16    3726.645   3808.739
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_SOME_BANNED.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1933.4231  
Iteration 2:   log pseudolikelihood = -1878.9284  (not concave)
Iteration 3:   log pseudolikelihood = -1794.8437  
Iteration 4:   log pseudolikelihood = -1775.7317  
Iteration 5:   log pseudolikelihood = -1774.4609  
Iteration 6:   log pseudolikelihood = -1774.4535  
Iteration 7:   log pseudolikelihood = -1774.4535  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     239.16
Log pseudolikelihood = -1774.454                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
        RESTRICT_SOME_BANNED |   .3757866   .0877567     4.28   0.000     .2037867    .5477865
      RESTRICTsq_SOME_BANNED |  -.0348427   .0108919    -3.20   0.001    -.0561905   -.0134949
                   PTS_Slag1 |   .4662305   .0891805     5.23   0.000       .29144    .6410211
                hrgroupslag1 |   .0001408   .0023591     0.06   0.952     -.004483    .0047646
                  hrnewslag1 |   .1288637   .0338801     3.80   0.000       .06246    .1952674
  protest_ClarkRegan_loglag1 |   .3014721   .0756257     3.99   0.000     .1532484    .4496958
      UCDP_armedConflictlag1 |    .265067   .2320723     1.14   0.253    -.1897863    .7199203
         PR_freedomHouselag1 |   .3705429   .2804457     1.32   0.186    -.1791205    .9202063
      PR_freedomHouselag1_sq |  -.0385135   .0324702    -1.19   0.236     -.102154    .0251269
   gdp_pc_constantUS2010lag1 |   .7306383   .3392782     2.15   0.031     .0656652    1.395611
gdp_pc_constantUS2010lag1_sq |  -.2740873   .1314012    -2.09   0.037    -.5316289   -.0165457
                   KOFGIlag1 |   .0870161   .0404269     2.15   0.031     .0077807    .1662515
                KOFGIlag1_sq |  -.0007955   .0003843    -2.07   0.038    -.0015487   -.0000423
              populationlag1 |   .0438241   .1807553     0.24   0.808    -.3104498    .3980981
                       _cons |  -3.934117   1.190876    -3.30   0.001    -6.268192   -1.600042
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .6038626   .5111172     1.18   0.237    -.3979088    1.605634
                     fhworst |  -1.720165   .9744711    -1.77   0.078    -3.630093    .1897635
      UCDP_armedConflictlag1 |  -.7821996   .4136781    -1.89   0.059    -1.592994    .0285945
                deathpenalty |   .1341399    .309617     0.43   0.665    -.4726982     .740978
           urgentActionslag1 |  -1.229031   .2267687    -5.42   0.000    -1.673489   -.7845723
                       _cons |   .3349762   .2567602     1.30   0.192    -.1682645    .8382169
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1203132   .1439305    -0.84   0.403    -.4024118    .1617854
-----------------------------+----------------------------------------------------------------
                       alpha |   .8866427   .1276149                      .6687053    1.175608
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1774.454      22    3592.907   3705.787
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_SOME_BANNED.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_SOME_BANNED |   .5778541   .1907242     3.03   0.002     .2040416    .9516665
      RESTRICTsq_SOME_BANNED |   -.051284   .0204031    -2.51   0.012    -.0912734   -.0112946
                   PTS_Slag1 |   .4850981   .1212067     4.00   0.000     .2475373    .7226588
                hrgroupslag1 |   -.000246   .0028603    -0.09   0.931    -.0058521    .0053602
                  hrnewslag1 |    .127093   .0282996     4.49   0.000     .0716269    .1825591
  protest_ClarkRegan_loglag1 |   .2720731   .1032605     2.63   0.008     .0696862      .47446
      UCDP_armedConflictlag1 |   .3523651   .2579784     1.37   0.172    -.1532632    .8579934
         PR_freedomHouselag1 |   .4877569   .3422712     1.43   0.154    -.1830823    1.158596
      PR_freedomHouselag1_sq |  -.0566132   .0390558    -1.45   0.147    -.1331612    .0199348
   gdp_pc_constantUS2010lag1 |   .9826255   .4901759     2.00   0.045     .0218984    1.943353
gdp_pc_constantUS2010lag1_sq |  -.4376192   .2440497    -1.79   0.073    -.9159478    .0407094
                   KOFGIlag1 |     .08672   .0808838     1.07   0.284    -.0718093    .2452493
                KOFGIlag1_sq |  -.0008686   .0007452    -1.17   0.244    -.0023291    .0005919
              populationlag1 |  -.0255108   .2498398    -0.10   0.919    -.5151877    .4641662
                       _cons |   -4.33059   2.097377    -2.06   0.039    -8.441373   -.2198078
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_SOME_BANNED RESTRICTsq_SOME_BANNED
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_SOME_BANNEDl2 RESTRICTsq_SOME_BANNEDl2
               RESTRICT_SOME_BANNEDl3
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) =  .15903 (p = 0.6901)

symmetric es_ic[1,1]
           c1
r1  .15902972
.\Tables\Appendix_TableS4_SOME_BANNED.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.x................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_SOME_BANNED |   .6041058   .2065028     2.93   0.003     .1993677    1.008844
      RESTRICTsq_SOME_BANNED |  -.0540752   .0209857    -2.58   0.010    -.0952064   -.0129439
                   PTS_Slag1 |    .634646   .1958991     3.24   0.001     .2506908    1.018601
         PR_freedomHouselag1 |   .3270927   .4136196     0.79   0.429    -.4835869    1.137772
      PR_freedomHouselag1_sq |  -.0426038    .047256    -0.90   0.367    -.1352238    .0500163
  protest_ClarkRegan_loglag1 |   .3637231   .1880656     1.93   0.053    -.0048787     .732325
   gdp_pc_constantUS2010lag1 |    .991704   .4898232     2.02   0.043     .0316682     1.95174
gdp_pc_constantUS2010lag1_sq |  -.4079821     .24566    -1.66   0.097    -.8894669    .0735027
                   KOFGIlag1 |    .082048   .0821037     1.00   0.318    -.0788722    .2429683
                KOFGIlag1_sq |  -.0007824   .0007663    -1.02   0.307    -.0022843    .0007195
                hrgroupslag1 |  -.0028672    .003835    -0.75   0.455    -.0103837    .0046492
                  hrnewslag1 |   .1095813   .0334982     3.27   0.001     .0439261    .1752365
              populationlag1 |  -.0131169   .2367242    -0.06   0.956    -.4770877     .450854
      UCDP_armedConflictlag1 |   .2391239   .3066624     0.78   0.436    -.3619234    .8401712
                       _cons |  -4.386525    2.25873    -1.94   0.052    -8.813555    .0405054
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_SOME_BANNED RESTRICTsq_SOME_BANNED PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_SOME_BANNEDl2 RESTRICTsq_SOME_BANNEDl2 RESTRICT_SOME_BANNEDl3
               PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
               protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .162138 (p = 0.6872)

symmetric es_ic[1,1]
           c1
r1  .16213815
.\Tables\Appendix_TableS4_SOME_BANNED.doc
dir : seeout
(2,400 missing values generated)
(2,576 missing values generated)
(2,576 missing values generated)
(2,576 missing values generated)
(2,576 missing values generated)
(2,576 missing values generated)
(2,576 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5119.5825  
Iteration 1:   log pseudolikelihood = -5119.0026  
Iteration 2:   log pseudolikelihood = -5119.0025  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3236.4265  
Iteration 1:   log pseudolikelihood = -2751.6549  
Iteration 2:   log pseudolikelihood = -2745.0628  
Iteration 3:   log pseudolikelihood = -2745.0556  
Iteration 4:   log pseudolikelihood = -2745.0556  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2638.5991  
Iteration 1:   log pseudolikelihood = -2617.9564  
Iteration 2:   log pseudolikelihood = -2592.2051  
Iteration 3:   log pseudolikelihood = -2592.1078  
Iteration 4:   log pseudolikelihood = -2592.1078  

Negative binomial regression                    Number of obs     =      1,690
                                                Wald chi2(2)      =      80.59
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2592.1078               Pseudo R2         =     0.0557

                                           (Std. Err. adjusted for 171 clusters in cowcode)
-------------------------------------------------------------------------------------------
                          |               Robust
            urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------------------+----------------------------------------------------------------
  RESTRICT_VISIT_RESTRICT |   .9514545   .1221505     7.79   0.000      .712044    1.190865
RESTRICTsq_VISIT_RESTRICT |  -.0826264    .012835    -6.44   0.000    -.1077826   -.0574703
                    _cons |  -.4126724   .1542791    -2.67   0.007    -.7150538    -.110291
--------------------------+----------------------------------------------------------------
                 /lnalpha |   1.167266   .0968213                      .9774994    1.357032
--------------------------+----------------------------------------------------------------
                    alpha |   3.213195   .3111058                      2.657802    3.884647
-------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,690 -2745.056  -2592.108       4    5192.216   5213.945
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1690  -2745.0556  -2592.1078           4   5192.2155   5213.9454
.\Tables\Appendix_TableS4_VISIT_RESTRICT.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -11997.909  
Iteration 1:   log pseudolikelihood = -6495.2999  
Iteration 2:   log pseudolikelihood = -4196.4284  
Iteration 3:   log pseudolikelihood = -2956.3721  
Iteration 4:   log pseudolikelihood =  -2865.445  
Iteration 5:   log pseudolikelihood = -2864.0468  
Iteration 6:   log pseudolikelihood = -2864.0459  
Iteration 7:   log pseudolikelihood = -2864.0459  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2007.7858  
Iteration 1:   log pseudolikelihood = -1849.9696  
Iteration 2:   log pseudolikelihood = -1837.5143  
Iteration 3:   log pseudolikelihood =  -1837.334  
Iteration 4:   log pseudolikelihood =  -1837.334  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     275.69
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood =  -1837.334               Pseudo R2         =     0.1411

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
     RESTRICT_VISIT_RESTRICT |   .5144757   .1063935     4.84   0.000     .3059483    .7230031
   RESTRICTsq_VISIT_RESTRICT |  -.0395538   .0110955    -3.56   0.000    -.0613006    -.017807
                   PTS_Slag1 |    .597552   .0989223     6.04   0.000     .4036678    .7914362
                hrgroupslag1 |   .0021444   .0027057     0.79   0.428    -.0031587    .0074474
                  hrnewslag1 |   .1803703   .0515528     3.50   0.000     .0793287    .2814119
  protest_ClarkRegan_loglag1 |   .3217778   .0919401     3.50   0.000     .1415785    .5019771
      UCDP_armedConflictlag1 |   .3543677    .240379     1.47   0.140    -.1167664    .8255019
         PR_freedomHouselag1 |   .5756174   .2263514     2.54   0.011     .1319767    1.019258
      PR_freedomHouselag1_sq |  -.0590922   .0271421    -2.18   0.029    -.1122899   -.0058946
   gdp_pc_constantUS2010lag1 |   .7385341   .3372248     2.19   0.029     .0775855    1.399483
gdp_pc_constantUS2010lag1_sq |  -.2567143   .1235171    -2.08   0.038    -.4988033   -.0146252
                   KOFGIlag1 |    .095861   .0442165     2.17   0.030     .0091983    .1825238
                KOFGIlag1_sq |  -.0008455   .0004209    -2.01   0.045    -.0016705   -.0000205
              populationlag1 |   .0898561   .1659018     0.54   0.588    -.2353054    .4150177
                       _cons |  -5.946649   1.255872    -4.74   0.000    -8.408113   -3.485185
-----------------------------+----------------------------------------------------------------
                    /lnalpha |    .487913   .1193341                      .2540225    .7218035
-----------------------------+----------------------------------------------------------------
                       alpha |   1.628913   .1943848                      1.289201    2.058142
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1837.334      16    3706.668   3788.762
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_VISIT_RESTRICT.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1813.4057  
Iteration 2:   log pseudolikelihood = -1770.5293  
Iteration 3:   log pseudolikelihood = -1765.8722  
Iteration 4:   log pseudolikelihood =  -1765.817  
Iteration 5:   log pseudolikelihood =  -1765.817  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     252.80
Log pseudolikelihood = -1765.817                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
     RESTRICT_VISIT_RESTRICT |   .4232201   .0894962     4.73   0.000     .2478109    .5986294
   RESTRICTsq_VISIT_RESTRICT |  -.0348126   .0096677    -3.60   0.000    -.0537609   -.0158642
                   PTS_Slag1 |   .4662823   .0920564     5.07   0.000     .2858551    .6467095
                hrgroupslag1 |   .0005079   .0024281     0.21   0.834    -.0042511    .0052668
                  hrnewslag1 |   .1208229   .0332304     3.64   0.000     .0556926    .1859532
  protest_ClarkRegan_loglag1 |   .3128189   .0752288     4.16   0.000     .1653731    .4602647
      UCDP_armedConflictlag1 |   .2382788   .2293424     1.04   0.299     -.211224    .6877815
         PR_freedomHouselag1 |   .3485856    .270756     1.29   0.198    -.1820863    .8792575
      PR_freedomHouselag1_sq |  -.0400479   .0310619    -1.29   0.197    -.1009282    .0208323
   gdp_pc_constantUS2010lag1 |   .6897606   .3347452     2.06   0.039     .0336721    1.345849
gdp_pc_constantUS2010lag1_sq |  -.2590095   .1281612    -2.02   0.043    -.5102008   -.0078182
                   KOFGIlag1 |    .079129   .0412143     1.92   0.055    -.0016496    .1599076
                KOFGIlag1_sq |  -.0007156   .0003916    -1.83   0.068    -.0014831    .0000518
              populationlag1 |   .0714462   .1745356     0.41   0.682    -.2706372    .4135297
                       _cons |  -3.785912   1.192306    -3.18   0.001    -6.122788   -1.449036
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .6477242   .5005328     1.29   0.196    -.3333021    1.628751
                     fhworst |  -1.800294   1.064452    -1.69   0.091    -3.886582    .2859935
      UCDP_armedConflictlag1 |  -.7865807   .4245111    -1.85   0.064    -1.618607    .0454458
                deathpenalty |   .1436017   .3169783     0.45   0.651    -.4776644    .7648679
           urgentActionslag1 |  -1.213538   .2254523    -5.38   0.000    -1.655416   -.7716596
                       _cons |   .2965555   .2685226     1.10   0.269     -.229739    .8228501
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1510197   .1441047    -1.05   0.295    -.4334597    .1314202
-----------------------------+----------------------------------------------------------------
                       alpha |   .8598307   .1239056                      .6482624    1.140447
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1765.817      22    3575.634   3688.514
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_VISIT_RESTRICT.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
     RESTRICT_VISIT_RESTRICT |   .6715279   .1597757     4.20   0.000     .3583733    .9846825
   RESTRICTsq_VISIT_RESTRICT |  -.0558148      .0169    -3.30   0.001    -.0889383   -.0226914
                   PTS_Slag1 |   .4852973   .1217953     3.98   0.000     .2465828    .7240118
                hrgroupslag1 |  -.0008179   .0029254    -0.28   0.780    -.0065515    .0049157
                  hrnewslag1 |   .1210863   .0276434     4.38   0.000     .0669063    .1752663
  protest_ClarkRegan_loglag1 |   .2903175    .096605     3.01   0.003     .1009751    .4796599
      UCDP_armedConflictlag1 |   .3122989   .2471299     1.26   0.206    -.1720667    .7966645
         PR_freedomHouselag1 |   .4295735   .3390541     1.27   0.205    -.2349604    1.094107
      PR_freedomHouselag1_sq |   -.052504   .0394253    -1.33   0.183    -.1297762    .0247682
   gdp_pc_constantUS2010lag1 |   .9475522   .4951791     1.91   0.056    -.0229809    1.918085
gdp_pc_constantUS2010lag1_sq |  -.4369222   .2608161    -1.68   0.094    -.9481123    .0742679
                   KOFGIlag1 |   .0775173   .0842958     0.92   0.358    -.0876994    .2427341
                KOFGIlag1_sq |   -.000765   .0007694    -0.99   0.320     -.002273    .0007431
              populationlag1 |   .0121176    .243652     0.05   0.960    -.4654315    .4896667
                       _cons |    -4.1501   2.166069    -1.92   0.055    -8.395516    .0953163
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_VISIT_RESTRICT RESTRICTsq_VISIT_RESTRICT
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_VISIT_RESTRICTl2 RESTRICTsq_VISIT_RESTRICTl2
               RESTRICT_VISIT_RESTRICTl3
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .054774 (p = 0.8150)

symmetric es_ic[1,1]
           c1
r1  .05477386
.\Tables\Appendix_TableS4_VISIT_RESTRICT.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
     RESTRICT_VISIT_RESTRICT |   .7181713   .1734445     4.14   0.000     .3782263    1.058116
   RESTRICTsq_VISIT_RESTRICT |  -.0614198   .0190991    -3.22   0.001    -.0988535   -.0239862
                   PTS_Slag1 |   .6303958   .1979277     3.18   0.001     .2424647    1.018327
         PR_freedomHouselag1 |   .1959314   .4350764     0.45   0.652    -.6568026    1.048665
      PR_freedomHouselag1_sq |  -.0289191   .0512062    -0.56   0.572    -.1292814    .0714433
  protest_ClarkRegan_loglag1 |   .3975701   .1648755     2.41   0.016       .07442    .7207202
   gdp_pc_constantUS2010lag1 |   .9751485    .508015     1.92   0.055    -.0205426     1.97084
gdp_pc_constantUS2010lag1_sq |  -.4109234   .2668913    -1.54   0.124    -.9340207    .1121739
                   KOFGIlag1 |   .0754522   .0862476     0.87   0.382      -.09359    .2444943
                KOFGIlag1_sq |  -.0006995   .0007987    -0.88   0.381    -.0022649    .0008658
                hrgroupslag1 |  -.0035903   .0038772    -0.93   0.354    -.0111894    .0040088
                  hrnewslag1 |   .1027034   .0329195     3.12   0.002     .0381824    .1672245
              populationlag1 |   .0316998   .2595531     0.12   0.903    -.4770149    .5404145
      UCDP_armedConflictlag1 |     .20347   .2851019     0.71   0.475    -.3553194    .7622595
                       _cons |  -4.188384   2.310279    -1.81   0.070    -8.716447     .339679
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_VISIT_RESTRICT RESTRICTsq_VISIT_RESTRICT PTS_Slag1
               PR_freedomHouselag1 PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_VISIT_RESTRICTl2 RESTRICTsq_VISIT_RESTRICTl2
               RESTRICT_VISIT_RESTRICTl3 PTS_Slag2 PR_freedomHouselag2
               PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .097966 (p = 0.7543)

symmetric es_ic[1,1]
           c1
r1  .09796604
.\Tables\Appendix_TableS4_VISIT_RESTRICT.doc
dir : seeout
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5273.5869  
Iteration 1:   log pseudolikelihood = -5273.4075  
Iteration 2:   log pseudolikelihood = -5273.4075  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2650.7279  
Iteration 1:   log pseudolikelihood = -2627.3278  
Iteration 2:   log pseudolikelihood = -2613.1221  
Iteration 3:   log pseudolikelihood =    -2613.1  
Iteration 4:   log pseudolikelihood =    -2613.1  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      61.04
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood =    -2613.1               Pseudo R2         =     0.0490

                                            (Std. Err. adjusted for 171 clusters in cowcode)
--------------------------------------------------------------------------------------------
                           |               Robust
             urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------------------+----------------------------------------------------------------
  RESTRICT_TRAVEL_RESTRICT |   .9002611   .1304421     6.90   0.000     .6445994    1.155923
RESTRICTsq_TRAVEL_RESTRICT |  -.0769729   .0133213    -5.78   0.000    -.1030822   -.0508637
                     _cons |  -.3729708   .1677971    -2.22   0.026     -.701847   -.0440946
---------------------------+----------------------------------------------------------------
                  /lnalpha |   1.213487   .0983153                      1.020792    1.406182
---------------------------+----------------------------------------------------------------
                     alpha |   3.365199   .3308507                      2.775393    4.080345
--------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809    -2613.1       4      5234.2   5255.932
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088     -2613.1           4   5234.1999   5255.9323
.\Tables\Appendix_TableS4_TRAVEL_RESTRICT.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12093.763  
Iteration 1:   log pseudolikelihood = -6477.7442  
Iteration 2:   log pseudolikelihood = -4230.5861  
Iteration 3:   log pseudolikelihood = -3009.0233  
Iteration 4:   log pseudolikelihood = -2905.0892  
Iteration 5:   log pseudolikelihood = -2903.4341  
Iteration 6:   log pseudolikelihood = -2903.4329  
Iteration 7:   log pseudolikelihood = -2903.4329  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2007.4973  
Iteration 1:   log pseudolikelihood = -1854.0791  
Iteration 2:   log pseudolikelihood =  -1841.758  
Iteration 3:   log pseudolikelihood = -1841.6625  
Iteration 4:   log pseudolikelihood = -1841.6625  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     284.03
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1841.6625               Pseudo R2         =     0.1391

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
    RESTRICT_TRAVEL_RESTRICT |   .4900483   .1076433     4.55   0.000     .2790714    .7010252
  RESTRICTsq_TRAVEL_RESTRICT |  -.0373644   .0110525    -3.38   0.001     -.059027   -.0157019
                   PTS_Slag1 |   .6211405   .1006043     6.17   0.000     .4239597    .8183212
                hrgroupslag1 |   .0020085   .0027188     0.74   0.460    -.0033202    .0073373
                  hrnewslag1 |   .1844979   .0544497     3.39   0.001     .0777785    .2912173
  protest_ClarkRegan_loglag1 |   .3242291   .0931278     3.48   0.000      .141702    .5067563
      UCDP_armedConflictlag1 |   .4084939   .2550111     1.60   0.109    -.0913186    .9083064
         PR_freedomHouselag1 |   .5716833   .2325239     2.46   0.014     .1159448    1.027422
      PR_freedomHouselag1_sq |  -.0584346   .0280503    -2.08   0.037    -.1134121   -.0034571
   gdp_pc_constantUS2010lag1 |   .7388236   .3306096     2.23   0.025     .0908407    1.386807
gdp_pc_constantUS2010lag1_sq |  -.2544674   .1232785    -2.06   0.039    -.4960889    -.012846
                   KOFGIlag1 |   .0930026   .0446743     2.08   0.037     .0054426    .1805625
                KOFGIlag1_sq |   -.000819   .0004222    -1.94   0.052    -.0016465    8.46e-06
              populationlag1 |   .0942083   .1696226     0.56   0.579    -.2382458    .4266624
                       _cons |  -5.925801   1.253421    -4.73   0.000    -8.382462    -3.46914
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5143975   .1196606                      .2798671    .7489278
-----------------------------+----------------------------------------------------------------
                       alpha |    1.67263   .2001479                      1.322954    2.114731
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1841.663      16    3715.325   3797.419
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_TRAVEL_RESTRICT.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1936.6766  
Iteration 2:   log pseudolikelihood =  -1885.282  (not concave)
Iteration 3:   log pseudolikelihood = -1800.7917  
Iteration 4:   log pseudolikelihood = -1774.9263  
Iteration 5:   log pseudolikelihood = -1771.8602  
Iteration 6:   log pseudolikelihood = -1771.6633  
Iteration 7:   log pseudolikelihood = -1771.6632  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     251.17
Log pseudolikelihood = -1771.663                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
    RESTRICT_TRAVEL_RESTRICT |   .3848443   .0912811     4.22   0.000     .2059367    .5637519
  RESTRICTsq_TRAVEL_RESTRICT |  -.0313364   .0097499    -3.21   0.001    -.0504458    -.012227
                   PTS_Slag1 |   .4871635    .093263     5.22   0.000     .3043714    .6699555
                hrgroupslag1 |   .0004352   .0024311     0.18   0.858    -.0043297    .0052001
                  hrnewslag1 |   .1235668   .0357034     3.46   0.001     .0535894    .1935442
  protest_ClarkRegan_loglag1 |   .3140122   .0762033     4.12   0.000     .1646565    .4633679
      UCDP_armedConflictlag1 |   .2765563   .2431841     1.14   0.255    -.2000758    .7531883
         PR_freedomHouselag1 |   .3671637   .2779006     1.32   0.186    -.1775115    .9118388
      PR_freedomHouselag1_sq |  -.0417077   .0322612    -1.29   0.196    -.1049386    .0215232
   gdp_pc_constantUS2010lag1 |   .6738181   .3336186     2.02   0.043     .0199376    1.327699
gdp_pc_constantUS2010lag1_sq |  -.2524901   .1277195    -1.98   0.048    -.5028157   -.0021646
                   KOFGIlag1 |   .0777689   .0420006     1.85   0.064    -.0045508    .1600887
                KOFGIlag1_sq |  -.0007023   .0003963    -1.77   0.076    -.0014791    .0000745
              populationlag1 |   .0712102   .1782059     0.40   0.689     -.278067    .4204873
                       _cons |  -3.832416   1.201281    -3.19   0.001    -6.186884   -1.477949
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .6132458   .5171255     1.19   0.236    -.4003016    1.626793
                     fhworst |  -1.791069   1.058357    -1.69   0.091    -3.865411    .2832726
      UCDP_armedConflictlag1 |  -.8227948    .443904    -1.85   0.064    -1.692831    .0472409
                deathpenalty |   .1486188   .3158186     0.47   0.638    -.4703743    .7676118
           urgentActionslag1 |  -1.238229   .2332807    -5.31   0.000    -1.695451   -.7810074
                       _cons |   .3008692   .2644495     1.14   0.255    -.2174424    .8191807
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   -.113004   .1459229    -0.77   0.439    -.3990077    .1729996
-----------------------------+----------------------------------------------------------------
                       alpha |   .8931471   .1303306                      .6709855    1.188866
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1771.663      22    3587.326   3700.206
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_TRAVEL_RESTRICT.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
    RESTRICT_TRAVEL_RESTRICT |   .5585367   .1518138     3.68   0.000     .2609872    .8560862
  RESTRICTsq_TRAVEL_RESTRICT |  -.0464389   .0169831    -2.73   0.006    -.0797251   -.0131526
                   PTS_Slag1 |   .5112459   .1265395     4.04   0.000      .263233    .7592587
                hrgroupslag1 |  -.0005579   .0030561    -0.18   0.855    -.0065478     .005432
                  hrnewslag1 |    .126789   .0283951     4.47   0.000     .0711355    .1824424
  protest_ClarkRegan_loglag1 |   .2928511   .1020536     2.87   0.004     .0928298    .4928725
      UCDP_armedConflictlag1 |   .3741715    .248114     1.51   0.132     -.112123     .860466
         PR_freedomHouselag1 |    .431203   .3527664     1.22   0.222    -.2602064    1.122612
      PR_freedomHouselag1_sq |  -.0511136   .0404993    -1.26   0.207    -.1304908    .0282637
   gdp_pc_constantUS2010lag1 |   .9221942   .5097535     1.81   0.070    -.0769043    1.921293
gdp_pc_constantUS2010lag1_sq |  -.4114369   .2567451    -1.60   0.109    -.9146481    .0917744
                   KOFGIlag1 |   .0847008   .0836531     1.01   0.311    -.0792563    .2486578
                KOFGIlag1_sq |   -.000831   .0007647    -1.09   0.277    -.0023298    .0006678
              populationlag1 |   .0065925   .2547526     0.03   0.979    -.4927135    .5058984
                       _cons |  -4.334434   2.102477    -2.06   0.039    -8.455214    -.213654
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_TRAVEL_RESTRICT RESTRICTsq_TRAVEL_RESTRICT
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_TRAVEL_RESTRICTl2 RESTRICTsq_TRAVEL_RESTRICTl2
               RESTRICT_TRAVEL_RESTRICTl3
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .004263 (p = 0.9479)

symmetric es_ic[1,1]
           c1
r1  .00426342
.\Tables\Appendix_TableS4_TRAVEL_RESTRICT.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.........................x........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
    RESTRICT_TRAVEL_RESTRICT |   .5942782   .1604784     3.70   0.000     .2797462    .9088101
  RESTRICTsq_TRAVEL_RESTRICT |  -.0504161   .0173625    -2.90   0.004     -.084446   -.0163861
                   PTS_Slag1 |    .648903   .2037532     3.18   0.001      .249554    1.048252
         PR_freedomHouselag1 |   .2541264    .432431     0.59   0.557    -.5934229    1.101676
      PR_freedomHouselag1_sq |  -.0340168    .049852    -0.68   0.495    -.1317248    .0636913
  protest_ClarkRegan_loglag1 |   .3890033   .1857181     2.09   0.036     .0250025    .7530042
   gdp_pc_constantUS2010lag1 |   .9293522   .5230696     1.78   0.076    -.0958453     1.95455
gdp_pc_constantUS2010lag1_sq |  -.3824078   .2633156    -1.45   0.146    -.8984969    .1336813
                   KOFGIlag1 |   .0798262   .0868675     0.92   0.358     -.090431    .2500834
                KOFGIlag1_sq |  -.0007409   .0008041    -0.92   0.357    -.0023168    .0008351
                hrgroupslag1 |  -.0030033   .0040166    -0.75   0.455    -.0108757    .0048691
                  hrnewslag1 |    .109751   .0348908     3.15   0.002     .0413662    .1781358
              populationlag1 |   .0246211   .2699665     0.09   0.927    -.5045036    .5537457
      UCDP_armedConflictlag1 |   .2673385   .2888156     0.93   0.355    -.2987296    .8334067
                       _cons |  -4.379651    2.25523    -1.94   0.052     -8.79982    .0405185
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_TRAVEL_RESTRICT RESTRICTsq_TRAVEL_RESTRICT PTS_Slag1
               PR_freedomHouselag1 PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_TRAVEL_RESTRICTl2 RESTRICTsq_TRAVEL_RESTRICTl2
               RESTRICT_TRAVEL_RESTRICTl3 PTS_Slag2 PR_freedomHouselag2
               PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .013508 (p = 0.9075)

symmetric es_ic[1,1]
           c1
r1  .01350759
.\Tables\Appendix_TableS4_TRAVEL_RESTRICT.doc
dir : seeout
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5128.8939  
Iteration 1:   log pseudolikelihood = -5128.5923  
Iteration 2:   log pseudolikelihood = -5128.5922  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2640.9012  
Iteration 1:   log pseudolikelihood = -2619.4891  
Iteration 2:   log pseudolikelihood = -2594.7505  
Iteration 3:   log pseudolikelihood =  -2594.658  
Iteration 4:   log pseudolikelihood = -2594.6579  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      78.52
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2594.6579               Pseudo R2         =     0.0557

                                        (Std. Err. adjusted for 171 clusters in cowcode)
----------------------------------------------------------------------------------------
                       |               Robust
         urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------+----------------------------------------------------------------
  RESTRICT_FUNDING_INT |   .9359563   .1203642     7.78   0.000     .7000469    1.171866
RESTRICTsq_FUNDING_INT |  -.0793447   .0123596    -6.42   0.000    -.1035691   -.0551203
                 _cons |  -.4617837   .1595406    -2.89   0.004    -.7744775     -.14909
-----------------------+----------------------------------------------------------------
              /lnalpha |   1.166316   .0983807                      .9734931    1.359138
-----------------------+----------------------------------------------------------------
                 alpha |   3.210144   .3158161                      2.647175    3.892837
----------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2594.658       4    5197.316   5219.048
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2594.6579           4   5197.3159   5219.0482
.\Tables\Appendix_TableS4_FUNDING_INT.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12059.248  
Iteration 1:   log pseudolikelihood = -6456.3837  
Iteration 2:   log pseudolikelihood = -4192.9646  
Iteration 3:   log pseudolikelihood = -2960.5012  
Iteration 4:   log pseudolikelihood = -2865.7758  
Iteration 5:   log pseudolikelihood =  -2864.456  
Iteration 6:   log pseudolikelihood = -2864.4552  
Iteration 7:   log pseudolikelihood = -2864.4552  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2008.6621  
Iteration 1:   log pseudolikelihood = -1852.6291  
Iteration 2:   log pseudolikelihood =  -1836.633  
Iteration 3:   log pseudolikelihood =  -1836.296  
Iteration 4:   log pseudolikelihood = -1836.2959  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     285.47
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1836.2959               Pseudo R2         =     0.1416

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_FUNDING_INT |   .5290574   .1059513     4.99   0.000     .3213966    .7367182
      RESTRICTsq_FUNDING_INT |  -.0401867   .0107524    -3.74   0.000     -.061261   -.0191124
                   PTS_Slag1 |   .5953959   .0961695     6.19   0.000     .4069072    .7838846
                hrgroupslag1 |   .0021448   .0027119     0.79   0.429    -.0031705      .00746
                  hrnewslag1 |   .1788829   .0524783     3.41   0.001     .0760273    .2817386
  protest_ClarkRegan_loglag1 |   .3257929   .0922294     3.53   0.000     .1450266    .5065591
      UCDP_armedConflictlag1 |   .3397875   .2467979     1.38   0.169    -.1439276    .8235026
         PR_freedomHouselag1 |   .5881956   .2277189     2.58   0.010     .1418747    1.034517
      PR_freedomHouselag1_sq |  -.0600348   .0274026    -2.19   0.028    -.1137429   -.0063268
   gdp_pc_constantUS2010lag1 |   .7078857   .3234965     2.19   0.029     .0738442    1.341927
gdp_pc_constantUS2010lag1_sq |   -.252272   .1219706    -2.07   0.039      -.49133   -.0132141
                   KOFGIlag1 |   .0970727   .0430896     2.25   0.024     .0126187    .1815267
                KOFGIlag1_sq |  -.0008332   .0004078    -2.04   0.041    -.0016325   -.0000338
              populationlag1 |   .0953746   .1646779     0.58   0.562    -.2273881    .4181372
                       _cons |  -6.130783    1.23317    -4.97   0.000    -8.547753   -3.713814
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .4910942   .1199874                      .2559233    .7262651
-----------------------------+----------------------------------------------------------------
                       alpha |   1.634103   .1960717                      1.291654    2.067345
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1836.296      16    3704.592   3786.686
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_FUNDING_INT.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1814.2887  
Iteration 2:   log pseudolikelihood = -1770.4571  
Iteration 3:   log pseudolikelihood = -1765.4447  
Iteration 4:   log pseudolikelihood = -1765.3883  
Iteration 5:   log pseudolikelihood = -1765.3883  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     245.77
Log pseudolikelihood = -1765.388                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
        RESTRICT_FUNDING_INT |   .4327451    .088311     4.90   0.000     .2596586    .6058315
      RESTRICTsq_FUNDING_INT |  -.0351099   .0092986    -3.78   0.000    -.0533349   -.0168849
                   PTS_Slag1 |   .4677498   .0887832     5.27   0.000      .293738    .6417617
                hrgroupslag1 |   .0005259    .002426     0.22   0.828     -.004229    .0052808
                  hrnewslag1 |   .1176166   .0327301     3.59   0.000     .0534668    .1817664
  protest_ClarkRegan_loglag1 |   .3157169   .0747172     4.23   0.000      .169274    .4621599
      UCDP_armedConflictlag1 |   .2272428   .2316013     0.98   0.327    -.2266875     .681173
         PR_freedomHouselag1 |   .3804622    .273604     1.39   0.164    -.1557917    .9167161
      PR_freedomHouselag1_sq |  -.0430494   .0314931    -1.37   0.172    -.1047746    .0186759
   gdp_pc_constantUS2010lag1 |   .6581258    .324498     2.03   0.043     .0221214     1.29413
gdp_pc_constantUS2010lag1_sq |  -.2544529   .1259681    -2.02   0.043    -.5013458     -.00756
                   KOFGIlag1 |    .080583   .0398085     2.02   0.043     .0025598    .1586063
                KOFGIlag1_sq |  -.0007108   .0003766    -1.89   0.059    -.0014489    .0000273
              populationlag1 |   .0682012   .1725402     0.40   0.693    -.2699715    .4063738
                       _cons |  -3.997173   1.165405    -3.43   0.001    -6.281326    -1.71302
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .6019153   .5141509     1.17   0.242     -.405802    1.609633
                     fhworst |   -1.84071   1.085576    -1.70   0.090    -3.968399    .2869797
      UCDP_armedConflictlag1 |  -.7722827   .4170916    -1.85   0.064    -1.589767    .0452018
                deathpenalty |   .1391795    .312102     0.45   0.656    -.4725292    .7508883
           urgentActionslag1 |  -1.197947   .2230009    -5.37   0.000    -1.635021   -.7608738
                       _cons |   .2994381   .2618946     1.14   0.253     -.213866    .8127422
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   -.151472   .1451835    -1.04   0.297    -.4360264    .1330824
-----------------------------+----------------------------------------------------------------
                       alpha |   .8594419   .1247768                      .6466007    1.142344
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1765.388      22    3574.777   3687.656
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_FUNDING_INT.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_FUNDING_INT |   .7395677   .1783778     4.15   0.000     .3899537    1.089182
      RESTRICTsq_FUNDING_INT |  -.0607342   .0176178    -3.45   0.001    -.0952646   -.0262039
                   PTS_Slag1 |    .504617   .1345999     3.75   0.000      .240806    .7684279
                hrgroupslag1 |  -.0009231   .0030568    -0.30   0.763    -.0069142    .0050681
                  hrnewslag1 |   .1136201   .0255594     4.45   0.000     .0635246    .1637156
  protest_ClarkRegan_loglag1 |   .3162618   .0988354     3.20   0.001      .122548    .5099755
      UCDP_armedConflictlag1 |   .2063397   .2604456     0.79   0.428    -.3041242    .7168037
         PR_freedomHouselag1 |   .3977155   .3487455     1.14   0.254    -.2858132    1.081244
      PR_freedomHouselag1_sq |  -.0501484   .0398856    -1.26   0.209    -.1283228    .0280259
   gdp_pc_constantUS2010lag1 |   1.036732   .4991695     2.08   0.038     .0583779    2.015086
gdp_pc_constantUS2010lag1_sq |  -.4814572   .2643819    -1.82   0.069    -.9996362    .0367217
                   KOFGIlag1 |   .0775704   .0812207     0.96   0.340    -.0816192    .2367599
                KOFGIlag1_sq |  -.0007695   .0007366    -1.04   0.296    -.0022132    .0006741
              populationlag1 |   .0079976   .2568925     0.03   0.975    -.4955025    .5114977
                       _cons |  -4.201101   2.132224    -1.97   0.049    -8.380184   -.0220183
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_FUNDING_INT RESTRICTsq_FUNDING_INT
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_FUNDING_INTl2 RESTRICTsq_FUNDING_INTl2
               RESTRICT_FUNDING_INTl3

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .067377 (p = 0.7952)

symmetric es_ic[1,1]
           c1
r1  .06737676
.\Tables\Appendix_TableS4_FUNDING_INT.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.x.....x..............x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_FUNDING_INT |   .7855934   .1797865     4.37   0.000     .4332182    1.137968
      RESTRICTsq_FUNDING_INT |   -.065072   .0177351    -3.67   0.000    -.0998321   -.0303119
                   PTS_Slag1 |   .6552335   .2145071     3.05   0.002     .2348073     1.07566
         PR_freedomHouselag1 |   .2199791   .4406637     0.50   0.618     -.643706    1.083664
      PR_freedomHouselag1_sq |  -.0349132   .0507824    -0.69   0.492    -.1344449    .0646184
  protest_ClarkRegan_loglag1 |   .3994152   .1670138     2.39   0.017     .0720741    .7267563
   gdp_pc_constantUS2010lag1 |   1.069656    .505523     2.12   0.034     .0788495    2.060463
gdp_pc_constantUS2010lag1_sq |  -.4632664   .2693944    -1.72   0.085    -.9912698     .064737
                   KOFGIlag1 |   .0724494   .0857759     0.84   0.398    -.0956683    .2405671
                KOFGIlag1_sq |  -.0006808   .0007877    -0.86   0.387    -.0022246     .000863
                hrgroupslag1 |  -.0036353   .0040273    -0.90   0.367    -.0115286     .004258
                  hrnewslag1 |   .0967466   .0308545     3.14   0.002      .036273    .1572202
              populationlag1 |    .022814   .2304589     0.10   0.921    -.4288771    .4745051
      UCDP_armedConflictlag1 |   .0795562   .3042818     0.26   0.794    -.5168251    .6759376
                       _cons |  -4.202855    2.34184    -1.79   0.073    -8.792776    .3870667
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_FUNDING_INT RESTRICTsq_FUNDING_INT PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_FUNDING_INTl2 RESTRICTsq_FUNDING_INTl2 RESTRICT_FUNDING_INTl3
               PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
               protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 3 bootstrap replicates;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .080756 (p = 0.7763)

symmetric es_ic[1,1]
          c1
r1  .0807563
.\Tables\Appendix_TableS4_FUNDING_INT.doc
dir : seeout
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5141.4626  
Iteration 1:   log pseudolikelihood = -5141.0547  
Iteration 2:   log pseudolikelihood = -5141.0546  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2641.0176  
Iteration 1:   log pseudolikelihood = -2620.1731  
Iteration 2:   log pseudolikelihood = -2594.5761  
Iteration 3:   log pseudolikelihood =  -2594.487  
Iteration 4:   log pseudolikelihood =  -2594.487  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      77.52
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood =  -2594.487               Pseudo R2         =     0.0558

                                        (Std. Err. adjusted for 171 clusters in cowcode)
----------------------------------------------------------------------------------------
                       |               Robust
         urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------+----------------------------------------------------------------
  RESTRICT_FUNDING_DOM |   .9396351   .1211708     7.75   0.000     .7021446    1.177126
RESTRICTsq_FUNDING_DOM |  -.0799664   .0124504    -6.42   0.000    -.1043688   -.0555641
                 _cons |  -.4616834   .1598989    -2.89   0.004    -.7750794   -.1482873
-----------------------+----------------------------------------------------------------
              /lnalpha |    1.16679   .0972063                      .9762687    1.357311
-----------------------+----------------------------------------------------------------
                 alpha |   3.211665   .3121943                      2.654533    3.885729
----------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2594.487       4    5196.974   5218.706
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088   -2594.487           4   5196.9739   5218.7062
.\Tables\Appendix_TableS4_FUNDING_DOM.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12062.596  
Iteration 1:   log pseudolikelihood = -6444.4143  
Iteration 2:   log pseudolikelihood = -4192.3733  
Iteration 3:   log pseudolikelihood = -2967.9988  
Iteration 4:   log pseudolikelihood = -2873.2522  
Iteration 5:   log pseudolikelihood = -2871.9853  
Iteration 6:   log pseudolikelihood = -2871.9847  
Iteration 7:   log pseudolikelihood = -2871.9847  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2007.7955  
Iteration 1:   log pseudolikelihood =  -1849.189  
Iteration 2:   log pseudolikelihood = -1836.4991  
Iteration 3:   log pseudolikelihood = -1836.3032  
Iteration 4:   log pseudolikelihood = -1836.3031  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     291.96
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1836.3031               Pseudo R2         =     0.1416

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_FUNDING_DOM |   .5260768    .104728     5.02   0.000     .3208138    .7313399
      RESTRICTsq_FUNDING_DOM |  -.0403271   .0107177    -3.76   0.000    -.0613334   -.0193209
                   PTS_Slag1 |   .5873592   .0960228     6.12   0.000      .399158    .7755604
                hrgroupslag1 |   .0021628   .0027065     0.80   0.424    -.0031418    .0074674
                  hrnewslag1 |   .1796972   .0522738     3.44   0.001     .0772424    .2821521
  protest_ClarkRegan_loglag1 |   .3301063    .091788     3.60   0.000     .1502052    .5100074
      UCDP_armedConflictlag1 |   .3550734   .2450157     1.45   0.147    -.1251485    .8352952
         PR_freedomHouselag1 |    .562417   .2279115     2.47   0.014     .1157186    1.009115
      PR_freedomHouselag1_sq |   -.057072   .0274109    -2.08   0.037    -.1107963   -.0033477
   gdp_pc_constantUS2010lag1 |   .7226523   .3216985     2.25   0.025     .0921348     1.35317
gdp_pc_constantUS2010lag1_sq |  -.2533357   .1215577    -2.08   0.037    -.4915845    -.015087
                   KOFGIlag1 |   .0999035   .0430576     2.32   0.020     .0155121    .1842949
                KOFGIlag1_sq |  -.0008719   .0004071    -2.14   0.032    -.0016699   -.0000739
              populationlag1 |   .0939598   .1650626     0.57   0.569     -.229557    .4174767
                       _cons |  -6.098837   1.230707    -4.96   0.000    -8.510979   -3.686695
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .4926777   .1195621                      .2583403    .7270151
-----------------------------+----------------------------------------------------------------
                       alpha |   1.636693   .1956865                      1.294779    2.068896
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1836.303      16    3704.606   3786.701
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_FUNDING_DOM.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1814.0339  
Iteration 2:   log pseudolikelihood = -1770.5421  
Iteration 3:   log pseudolikelihood = -1765.6481  
Iteration 4:   log pseudolikelihood = -1765.5921  
Iteration 5:   log pseudolikelihood =  -1765.592  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     254.94
Log pseudolikelihood = -1765.592                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
        RESTRICT_FUNDING_DOM |    .428651   .0879214     4.88   0.000     .2563282    .6009738
      RESTRICTsq_FUNDING_DOM |  -.0351014   .0093319    -3.76   0.000    -.0533917   -.0168111
                   PTS_Slag1 |   .4620133   .0883959     5.23   0.000     .2887606     .635266
                hrgroupslag1 |   .0005063   .0024114     0.21   0.834    -.0042199    .0052326
                  hrnewslag1 |    .118493   .0329334     3.60   0.000     .0539448    .1830413
  protest_ClarkRegan_loglag1 |   .3204571   .0742061     4.32   0.000     .1750159    .4658983
      UCDP_armedConflictlag1 |   .2383073   .2316652     1.03   0.304    -.2157482    .6923628
         PR_freedomHouselag1 |   .3499003     .27311     1.28   0.200    -.1853855     .885186
      PR_freedomHouselag1_sq |  -.0395012   .0314471    -1.26   0.209    -.1011364     .022134
   gdp_pc_constantUS2010lag1 |   .6654245   .3239828     2.05   0.040     .0304298    1.300419
gdp_pc_constantUS2010lag1_sq |  -.2532523   .1258742    -2.01   0.044    -.4999612   -.0065433
                   KOFGIlag1 |   .0833855   .0397558     2.10   0.036     .0054655    .1613055
                KOFGIlag1_sq |  -.0007456   .0003752    -1.99   0.047    -.0014809   -.0000102
              populationlag1 |   .0695296   .1724951     0.40   0.687    -.2685545    .4076138
                       _cons |  -3.974272   1.163655    -3.42   0.001    -6.254994   -1.693551
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .6076272   .5132473     1.18   0.236     -.398319    1.613573
                     fhworst |  -1.786324   1.030794    -1.73   0.083    -3.806644     .233996
      UCDP_armedConflictlag1 |  -.7920532   .4292644    -1.85   0.065    -1.633396    .0492896
                deathpenalty |   .1352166   .3137443     0.43   0.666    -.4797109    .7501442
           urgentActionslag1 |  -1.214758    .230296    -5.27   0.000     -1.66613   -.7633858
                       _cons |   .3013889    .261629     1.15   0.249    -.2113944    .8141722
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1454453   .1443967    -1.01   0.314    -.4284577     .137567
-----------------------------+----------------------------------------------------------------
                       alpha |   .8646372   .1248508                      .6515132    1.147479
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1765.592      22    3575.184   3688.064
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_FUNDING_DOM.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_FUNDING_DOM |   .7047629   .1592822     4.42   0.000     .3925755     1.01695
      RESTRICTsq_FUNDING_DOM |  -.0582128   .0167855    -3.47   0.001    -.0911117   -.0253139
                   PTS_Slag1 |   .4868041   .1254103     3.88   0.000     .2410044    .7326037
                hrgroupslag1 |  -.0009873   .0030353    -0.33   0.745    -.0069363    .0049617
                  hrnewslag1 |   .1168929   .0276491     4.23   0.000     .0627018    .1710841
  protest_ClarkRegan_loglag1 |   .3065142   .0985718     3.11   0.002      .113317    .4997114
      UCDP_armedConflictlag1 |   .2827713   .2403398     1.18   0.239     -.188286    .7538286
         PR_freedomHouselag1 |   .3696179   .3410538     1.08   0.278    -.2988352    1.038071
      PR_freedomHouselag1_sq |  -.0463825   .0395028    -1.17   0.240    -.1238065    .0310415
   gdp_pc_constantUS2010lag1 |   1.006898   .4994779     2.02   0.044     .0279389    1.985856
gdp_pc_constantUS2010lag1_sq |  -.4596334   .2604197    -1.76   0.078    -.9700466    .0507797
                   KOFGIlag1 |    .079452   .0818243     0.97   0.332    -.0809207    .2398246
                KOFGIlag1_sq |  -.0007841    .000741    -1.06   0.290    -.0022364    .0006682
              populationlag1 |   .0113316   .2422461     0.05   0.963     -.463462    .4861252
                       _cons |  -4.139857   2.139805    -1.93   0.053    -8.333798    .0540837
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_FUNDING_DOM RESTRICTsq_FUNDING_DOM
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_FUNDING_DOMl2 RESTRICTsq_FUNDING_DOMl2
               RESTRICT_FUNDING_DOMl3
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .060343 (p = 0.8060)

symmetric es_ic[1,1]
           c1
r1  .06034302
.\Tables\Appendix_TableS4_FUNDING_DOM.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.x................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
        RESTRICT_FUNDING_DOM |   .7504414   .1762615     4.26   0.000     .4049751    1.095908
      RESTRICTsq_FUNDING_DOM |  -.0625774   .0217006    -2.88   0.004    -.1051098   -.0200449
                   PTS_Slag1 |   .6352836   .2009251     3.16   0.002     .2414776     1.02909
         PR_freedomHouselag1 |    .162443   .4375375     0.37   0.710    -.6951149    1.020001
      PR_freedomHouselag1_sq |  -.0280881   .0514962    -0.55   0.585    -.1290188    .0728426
  protest_ClarkRegan_loglag1 |   .3714183   .1692634     2.19   0.028     .0396682    .7031684
   gdp_pc_constantUS2010lag1 |   1.019704   .4986461     2.04   0.041      .042376    1.997033
gdp_pc_constantUS2010lag1_sq |  -.4334039   .2629909    -1.65   0.099    -.9488567    .0820488
                   KOFGIlag1 |   .0759187   .0847024     0.90   0.370     -.090095    .2419323
                KOFGIlag1_sq |  -.0007124    .000776    -0.92   0.359    -.0022334    .0008086
                hrgroupslag1 |  -.0036458   .0040405    -0.90   0.367    -.0115651    .0042734
                  hrnewslag1 |   .1013805   .0332065     3.05   0.002     .0362969    .1664641
              populationlag1 |   .0224287   .2294726     0.10   0.922    -.4273294    .4721867
      UCDP_armedConflictlag1 |   .1657929   .2806939     0.59   0.555     -.384357    .7159429
                       _cons |  -4.100762    2.31008    -1.78   0.076    -8.628436    .4269121
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_FUNDING_DOM RESTRICTsq_FUNDING_DOM PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_FUNDING_DOMl2 RESTRICTsq_FUNDING_DOMl2 RESTRICT_FUNDING_DOMl3
               PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
               protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .085305 (p = 0.7702)

symmetric es_ic[1,1]
           c1
r1  .08530499
.\Tables\Appendix_TableS4_FUNDING_DOM.doc
dir : seeout
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5137.1123  
Iteration 1:   log pseudolikelihood = -5136.7358  
Iteration 2:   log pseudolikelihood = -5136.7357  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2645.1805  
Iteration 1:   log pseudolikelihood = -2622.1488  
Iteration 2:   log pseudolikelihood = -2603.1225  
Iteration 3:   log pseudolikelihood = -2603.0529  
Iteration 4:   log pseudolikelihood = -2603.0529  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      73.49
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2603.0529               Pseudo R2         =     0.0527

                                         (Std. Err. adjusted for 171 clusters in cowcode)
-----------------------------------------------------------------------------------------
                        |               Robust
          urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------+----------------------------------------------------------------
  RESTRICT_REGISTRATION |   .9196889   .1177907     7.81   0.000     .6888233    1.150554
RESTRICTsq_REGISTRATION |  -.0811574   .0126483    -6.42   0.000    -.1059477   -.0563672
                  _cons |   -.366421   .1639745    -2.23   0.025    -.6878051   -.0450369
------------------------+----------------------------------------------------------------
               /lnalpha |   1.181403   .0990692                      .9872306    1.375575
------------------------+----------------------------------------------------------------
                  alpha |   3.258942   .3228607                      2.683792     3.95735
-----------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2603.053       4    5214.106   5235.838
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2603.0529           4   5214.1057    5235.838
.\Tables\Appendix_TableS4_REGISTRATION.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12091.779  
Iteration 1:   log pseudolikelihood = -6508.3845  
Iteration 2:   log pseudolikelihood =  -4242.159  
Iteration 3:   log pseudolikelihood = -3002.7204  
Iteration 4:   log pseudolikelihood = -2907.5976  
Iteration 5:   log pseudolikelihood = -2906.1067  
Iteration 6:   log pseudolikelihood = -2906.1057  
Iteration 7:   log pseudolikelihood = -2906.1057  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2007.9747  
Iteration 1:   log pseudolikelihood = -1854.1931  
Iteration 2:   log pseudolikelihood = -1842.2208  
Iteration 3:   log pseudolikelihood = -1842.1247  
Iteration 4:   log pseudolikelihood = -1842.1247  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     278.01
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1842.1247               Pseudo R2         =     0.1389

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
       RESTRICT_REGISTRATION |   .5127268   .1106554     4.63   0.000     .2958462    .7296073
     RESTRICTsq_REGISTRATION |  -.0410378   .0119836    -3.42   0.001    -.0645253   -.0175504
                   PTS_Slag1 |   .5792235    .096949     5.97   0.000     .3892071      .76924
                hrgroupslag1 |   .0018362   .0026718     0.69   0.492    -.0034005    .0070728
                  hrnewslag1 |   .1770178   .0517831     3.42   0.001     .0755248    .2785108
  protest_ClarkRegan_loglag1 |   .3230457   .0936404     3.45   0.001      .139514    .5065775
      UCDP_armedConflictlag1 |   .3526192   .2477325     1.42   0.155    -.1329277    .8381661
         PR_freedomHouselag1 |   .6380062   .2278135     2.80   0.005     .1914999    1.084513
      PR_freedomHouselag1_sq |  -.0643606   .0274152    -2.35   0.019    -.1180935   -.0106277
   gdp_pc_constantUS2010lag1 |   .7532844   .3267837     2.31   0.021        .1128    1.393769
gdp_pc_constantUS2010lag1_sq |  -.2617639   .1239713    -2.11   0.035    -.5047433   -.0187846
                   KOFGIlag1 |    .098124    .044091     2.23   0.026     .0117072    .1845408
                KOFGIlag1_sq |  -.0008524   .0004179    -2.04   0.041    -.0016714   -.0000333
              populationlag1 |   .1127143   .1698282     0.66   0.507    -.2201429    .4455715
                       _cons |  -6.083042   1.271126    -4.79   0.000    -8.574403    -3.59168
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5152261   .1195983                      .2808176    .7496345
-----------------------------+----------------------------------------------------------------
                       alpha |   1.674017   .2002097                      1.324212    2.116226
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1842.125      16    3716.249   3798.344
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_REGISTRATION.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1817.7439  
Iteration 2:   log pseudolikelihood = -1775.7226  
Iteration 3:   log pseudolikelihood = -1771.3344  
Iteration 4:   log pseudolikelihood = -1771.2701  
Iteration 5:   log pseudolikelihood =   -1771.27  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     236.57
Log pseudolikelihood =  -1771.27                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
       RESTRICT_REGISTRATION |   .4045268   .0926943     4.36   0.000     .2228493    .5862043
     RESTRICTsq_REGISTRATION |   -.034583   .0104893    -3.30   0.001    -.0551417   -.0140243
                   PTS_Slag1 |   .4475187   .0892035     5.02   0.000      .272683    .6223544
                hrgroupslag1 |    .000285   .0023664     0.12   0.904    -.0043531    .0049231
                  hrnewslag1 |     .11747   .0335691     3.50   0.000     .0516757    .1832642
  protest_ClarkRegan_loglag1 |   .3121816   .0763498     4.09   0.000     .1625387    .4618246
      UCDP_armedConflictlag1 |   .2504406   .2329613     1.08   0.282    -.2061551    .7070363
         PR_freedomHouselag1 |   .4350122   .2707019     1.61   0.108    -.0955537    .9655781
      PR_freedomHouselag1_sq |  -.0480744   .0311352    -1.54   0.123    -.1090983    .0129495
   gdp_pc_constantUS2010lag1 |   .7060758   .3345246     2.11   0.035     .0504195    1.361732
gdp_pc_constantUS2010lag1_sq |  -.2671545   .1307952    -2.04   0.041    -.5235084   -.0108006
                   KOFGIlag1 |   .0842701    .040691     2.07   0.038     .0045172     .164023
                KOFGIlag1_sq |  -.0007557    .000386    -1.96   0.050    -.0015123    9.40e-07
              populationlag1 |   .0897043   .1772983     0.51   0.613     -.257794    .4372027
                       _cons |  -4.008401    1.20273    -3.33   0.001    -6.365708   -1.651093
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .5770502   .5214076     1.11   0.268    -.4448899     1.59899
                     fhworst |   -1.82265   1.056565    -1.73   0.085     -3.89348    .2481804
      UCDP_armedConflictlag1 |  -.7864521   .4238464    -1.86   0.064    -1.617176    .0442716
                deathpenalty |   .1624947   .3142221     0.52   0.605    -.4533694    .7783587
           urgentActionslag1 |  -1.229868   .2302767    -5.34   0.000    -1.681202   -.7785338
                       _cons |   .3026994   .2643895     1.14   0.252    -.2154945    .8208933
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1215466   .1418195    -0.86   0.391    -.3995077    .1564145
-----------------------------+----------------------------------------------------------------
                       alpha |   .8855498   .1255882                      .6706501    1.169311
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85   -1771.27      22     3586.54    3699.42
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_REGISTRATION.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
       RESTRICT_REGISTRATION |    .702923     .17676     3.98   0.000     .3564797    1.049366
     RESTRICTsq_REGISTRATION |  -.0613582   .0192849    -3.18   0.001    -.0991559   -.0235604
                   PTS_Slag1 |   .4685462   .1233116     3.80   0.000     .2268598    .7102325
                hrgroupslag1 |  -.0012674   .0030303    -0.42   0.676    -.0072067     .004672
                  hrnewslag1 |    .115776   .0267748     4.32   0.000     .0632984    .1682536
  protest_ClarkRegan_loglag1 |   .2856383   .0993685     2.87   0.004     .0908796     .480397
      UCDP_armedConflictlag1 |   .2816974   .2469595     1.14   0.254    -.2023343    .7657291
         PR_freedomHouselag1 |   .4192981   .3397487     1.23   0.217    -.2465972    1.085193
      PR_freedomHouselag1_sq |  -.0511801    .038763    -1.32   0.187    -.1271542     .024794
   gdp_pc_constantUS2010lag1 |   1.011719   .5104559     1.98   0.047     .0112444    2.012195
gdp_pc_constantUS2010lag1_sq |  -.4344158   .2531697    -1.72   0.086    -.9306193    .0617877
                   KOFGIlag1 |   .0878902   .0813308     1.08   0.280    -.0715153    .2472956
                KOFGIlag1_sq |  -.0008675   .0007456    -1.16   0.245    -.0023288    .0005938
              populationlag1 |    .034826   .2505253     0.14   0.889    -.4561946    .5258466
                       _cons |  -4.275804   2.131238    -2.01   0.045    -8.452953   -.0986544
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_REGISTRATION RESTRICTsq_REGISTRATION
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_REGISTRATIONl2 RESTRICTsq_REGISTRATIONl2
               RESTRICT_REGISTRATIONl3
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) =  .09618 (p = 0.7565)

symmetric es_ic[1,1]
           c1
r1  .09617982
.\Tables\Appendix_TableS4_REGISTRATION.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
       RESTRICT_REGISTRATION |   .7435029   .1960874     3.79   0.000     .3591786    1.127827
     RESTRICTsq_REGISTRATION |  -.0655605   .0202325    -3.24   0.001    -.1052154   -.0259055
                   PTS_Slag1 |   .6205159   .1958301     3.17   0.002      .236696    1.004336
         PR_freedomHouselag1 |    .222363   .4139736     0.54   0.591    -.5890102    1.033736
      PR_freedomHouselag1_sq |  -.0337338   .0465851    -0.72   0.469    -.1250389    .0575713
  protest_ClarkRegan_loglag1 |   .3649539   .1776887     2.05   0.040     .0166905    .7132173
   gdp_pc_constantUS2010lag1 |   1.029462   .5177449     1.99   0.047     .0147008    2.044224
gdp_pc_constantUS2010lag1_sq |  -.4070325   .2579703    -1.58   0.115     -.912645      .09858
                   KOFGIlag1 |   .0848513   .0823728     1.03   0.303    -.0765964    .2462991
                KOFGIlag1_sq |  -.0007984   .0007655    -1.04   0.297    -.0022987    .0007019
                hrgroupslag1 |  -.0040245   .0039052    -1.03   0.303    -.0116786    .0036295
                  hrnewslag1 |   .0986599   .0309589     3.19   0.001     .0379816    .1593381
              populationlag1 |   .0497647   .2655624     0.19   0.851    -.4707281    .5702575
      UCDP_armedConflictlag1 |    .163277   .2926609     0.56   0.577    -.4103279    .7368819
                       _cons |  -4.287631   2.277619    -1.88   0.060    -8.751682     .176419
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_REGISTRATION RESTRICTsq_REGISTRATION PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_REGISTRATIONl2 RESTRICTsq_REGISTRATIONl2 RESTRICT_REGISTRATIONl3
               PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
               protest_ClarkRegan_loglag2

  Test of overidentifying restriction:

  Hansen's J chi2(1) =  .12505 (p = 0.7236)

symmetric es_ic[1,1]
           c1
r1  .12504998
.\Tables\Appendix_TableS4_REGISTRATION.doc
dir : seeout
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5108.5356  
Iteration 1:   log pseudolikelihood = -5108.2349  
Iteration 2:   log pseudolikelihood = -5108.2348  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2640.2571  
Iteration 1:   log pseudolikelihood =  -2619.251  
Iteration 2:   log pseudolikelihood =  -2593.593  
Iteration 3:   log pseudolikelihood = -2593.4889  
Iteration 4:   log pseudolikelihood = -2593.4889  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      78.12
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2593.4889               Pseudo R2         =     0.0562

                                   (Std. Err. adjusted for 171 clusters in cowcode)
-----------------------------------------------------------------------------------
                  |               Robust
    urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------+----------------------------------------------------------------
  RESTRICT_CENSOR |   .9477553   .1202671     7.88   0.000     .7120362    1.183474
RESTRICTsq_CENSOR |  -.0825296    .012785    -6.46   0.000    -.1075877   -.0574715
            _cons |  -.4566713   .1574705    -2.90   0.004    -.7653077   -.1480349
------------------+----------------------------------------------------------------
         /lnalpha |   1.162778   .0976809                       .971327    1.354229
------------------+----------------------------------------------------------------
            alpha |   3.198808   .3124624                      2.641447    3.873774
-----------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2593.489       4    5194.978    5216.71
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2593.4889           4   5194.9777     5216.71
.\Tables\Appendix_TableS4_CENSOR.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -11987.952  
Iteration 1:   log pseudolikelihood = -6538.1053  
Iteration 2:   log pseudolikelihood = -4218.7983  
Iteration 3:   log pseudolikelihood = -2964.9466  
Iteration 4:   log pseudolikelihood = -2874.2595  
Iteration 5:   log pseudolikelihood = -2872.6857  
Iteration 6:   log pseudolikelihood = -2872.6846  
Iteration 7:   log pseudolikelihood = -2872.6846  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2008.0711  
Iteration 1:   log pseudolikelihood = -1850.2271  
Iteration 2:   log pseudolikelihood = -1837.6665  
Iteration 3:   log pseudolikelihood = -1837.4827  
Iteration 4:   log pseudolikelihood = -1837.4827  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     288.44
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1837.4827               Pseudo R2         =     0.1410

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
             RESTRICT_CENSOR |   .5303227   .1098708     4.83   0.000       .31498    .7456655
           RESTRICTsq_CENSOR |  -.0416121   .0116921    -3.56   0.000    -.0645281   -.0186961
                   PTS_Slag1 |   .5966068   .0959702     6.22   0.000     .4085085     .784705
                hrgroupslag1 |   .0022548   .0026987     0.84   0.403    -.0030345    .0075441
                  hrnewslag1 |   .1758312   .0509089     3.45   0.001     .0760515    .2756108
  protest_ClarkRegan_loglag1 |   .3129798   .0935935     3.34   0.001     .1295399    .4964197
      UCDP_armedConflictlag1 |   .3224387   .2464397     1.31   0.191    -.1605742    .8054515
         PR_freedomHouselag1 |   .6155563   .2264775     2.72   0.007     .1716686    1.059444
      PR_freedomHouselag1_sq |   -.063037   .0271925    -2.32   0.020    -.1163332   -.0097407
   gdp_pc_constantUS2010lag1 |   .7386352   .3324677     2.22   0.026     .0870104     1.39026
gdp_pc_constantUS2010lag1_sq |  -.2587903   .1235996    -2.09   0.036    -.5010411   -.0165394
                   KOFGIlag1 |   .0987509   .0430206     2.30   0.022     .0144322    .1830696
                KOFGIlag1_sq |  -.0008519   .0004093    -2.08   0.037    -.0016541   -.0000497
              populationlag1 |    .106238   .1642629     0.65   0.518    -.2157114    .4281873
                       _cons |  -6.183496   1.228467    -5.03   0.000    -8.591247   -3.775744
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .4958972   .1197828                      .2611272    .7306672
-----------------------------+----------------------------------------------------------------
                       alpha |   1.641971   .1966799                      1.298393    2.076466
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1837.483      16    3706.965    3789.06
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_CENSOR.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1814.2481  
Iteration 2:   log pseudolikelihood = -1770.9457  
Iteration 3:   log pseudolikelihood = -1766.1219  
Iteration 4:   log pseudolikelihood = -1766.0691  
Iteration 5:   log pseudolikelihood = -1766.0691  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     252.29
Log pseudolikelihood = -1766.069                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
             RESTRICT_CENSOR |   .4334435   .0932383     4.65   0.000     .2506997    .6161872
           RESTRICTsq_CENSOR |  -.0362396   .0103304    -3.51   0.000    -.0564869   -.0159924
                   PTS_Slag1 |   .4696116   .0885699     5.30   0.000     .2960178    .6432054
                hrgroupslag1 |   .0006675   .0024057     0.28   0.781    -.0040476    .0053825
                  hrnewslag1 |   .1166668   .0323921     3.60   0.000     .0531794    .1801542
  protest_ClarkRegan_loglag1 |   .3031871   .0755722     4.01   0.000     .1550682    .4513059
      UCDP_armedConflictlag1 |   .2112838   .2317538     0.91   0.362    -.2429453    .6655129
         PR_freedomHouselag1 |   .4087109   .2705003     1.51   0.131      -.12146    .9388818
      PR_freedomHouselag1_sq |  -.0463182    .031075    -1.49   0.136    -.1072242    .0145877
   gdp_pc_constantUS2010lag1 |   .6938333    .333537     2.08   0.038     .0401127    1.347554
gdp_pc_constantUS2010lag1_sq |  -.2626686    .128397    -2.05   0.041    -.5143221   -.0110151
                   KOFGIlag1 |   .0827265   .0397908     2.08   0.038      .004738    .1607151
                KOFGIlag1_sq |  -.0007345   .0003783    -1.94   0.052     -.001476    7.08e-06
              populationlag1 |   .0856418   .1697175     0.50   0.614    -.2469983    .4182819
                       _cons |  -4.065948   1.161951    -3.50   0.000     -6.34333   -1.788567
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |    .582628   .5163011     1.13   0.259    -.4293035    1.594559
                     fhworst |  -1.792192   1.024123    -1.75   0.080    -3.799437    .2150522
      UCDP_armedConflictlag1 |  -.7749372   .4149577    -1.87   0.062    -1.588239    .0383649
                deathpenalty |   .1402277   .3124321     0.45   0.654     -.472128    .7525834
           urgentActionslag1 |  -1.200964   .2261395    -5.31   0.000    -1.644189   -.7577384
                       _cons |   .3048639   .2607701     1.17   0.242    -.2062362     .815964
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1504952   .1458439    -1.03   0.302     -.436344    .1353536
-----------------------------+----------------------------------------------------------------
                       alpha |   .8602818   .1254669                      .6463953    1.144942
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1766.069      22    3576.138   3689.018
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_CENSOR.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x..x........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
             RESTRICT_CENSOR |   .7249796    .178517     4.06   0.000     .3750928    1.074866
           RESTRICTsq_CENSOR |  -.0610554   .0186435    -3.27   0.001    -.0975959   -.0245149
                   PTS_Slag1 |   .4971112   .1257858     3.95   0.000     .2505756    .7436469
                hrgroupslag1 |   -.000556   .0030349    -0.18   0.855    -.0065042    .0053922
                  hrnewslag1 |   .1166848   .0263439     4.43   0.000     .0650518    .1683178
  protest_ClarkRegan_loglag1 |   .2853014   .1002312     2.85   0.004     .0888518     .481751
      UCDP_armedConflictlag1 |   .2272517   .2574793     0.88   0.377    -.2773986    .7319019
         PR_freedomHouselag1 |   .4036743   .3453519     1.17   0.242    -.2732031    1.080552
      PR_freedomHouselag1_sq |  -.0515312   .0394774    -1.31   0.192    -.1289054     .025843
   gdp_pc_constantUS2010lag1 |   .9969786   .5064067     1.97   0.049     .0044396    1.989518
gdp_pc_constantUS2010lag1_sq |   -.456069   .2595696    -1.76   0.079     -.964816    .0526781
                   KOFGIlag1 |   .0777716   .0813088     0.96   0.339    -.0815908    .2371339
                KOFGIlag1_sq |  -.0007768   .0007417    -1.05   0.295    -.0022305     .000677
              populationlag1 |   .0162624   .2434967     0.07   0.947    -.4609823    .4935071
                       _cons |  -4.147934   2.122548    -1.95   0.051    -8.308052    .0121844
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_CENSOR RESTRICTsq_CENSOR
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_CENSORl2 RESTRICTsq_CENSORl2 RESTRICT_CENSORl3
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .005379 (p = 0.9415)

symmetric es_ic[1,1]
           c1
r1  .00537875
.\Tables\Appendix_TableS4_CENSOR.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.......x..........................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
             RESTRICT_CENSOR |   .7677885   .1944122     3.95   0.000     .3867477    1.148829
           RESTRICTsq_CENSOR |  -.0655392   .0195139    -3.36   0.001    -.1037856   -.0272927
                   PTS_Slag1 |   .6623242   .2011952     3.29   0.001     .2679888     1.05666
         PR_freedomHouselag1 |   .1961519   .4256869     0.46   0.645    -.6381792    1.030483
      PR_freedomHouselag1_sq |  -.0332023   .0481245    -0.69   0.490    -.1275245    .0611199
  protest_ClarkRegan_loglag1 |   .3418603   .1856173     1.84   0.066     -.021943    .7056636
   gdp_pc_constantUS2010lag1 |   1.012843   .5135043     1.97   0.049     .0063935    2.019293
gdp_pc_constantUS2010lag1_sq |  -.4289759   .2632904    -1.63   0.103    -.9450155    .0870637
                   KOFGIlag1 |   .0744223   .0830366     0.90   0.370    -.0883265    .2371712
                KOFGIlag1_sq |  -.0007046   .0007675    -0.92   0.359    -.0022089    .0007996
                hrgroupslag1 |  -.0031823   .0039218    -0.81   0.417    -.0108689    .0045044
                  hrnewslag1 |    .101014   .0315599     3.20   0.001     .0391578    .1628702
              populationlag1 |   .0240528   .2572971     0.09   0.926    -.4802403    .5283459
      UCDP_armedConflictlag1 |   .0920529   .3043666     0.30   0.762    -.5044947    .6886005
                       _cons |  -4.149364   2.267273    -1.83   0.067    -8.593137    .2944082
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_CENSOR RESTRICTsq_CENSOR PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_CENSORl2 RESTRICTsq_CENSORl2 RESTRICT_CENSORl3 PTS_Slag2
               PR_freedomHouselag2 PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .015123 (p = 0.9021)

symmetric es_ic[1,1]
           c1
r1  .01512338
.\Tables\Appendix_TableS4_CENSOR.doc
dir : seeout
(2,399 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5436.4992  
Iteration 1:   log pseudolikelihood = -5435.5874  
Iteration 2:   log pseudolikelihood = -5435.5872  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2664.5523  
Iteration 1:   log pseudolikelihood = -2643.7498  
Iteration 2:   log pseudolikelihood = -2636.8283  
Iteration 3:   log pseudolikelihood = -2636.8236  
Iteration 4:   log pseudolikelihood = -2636.8236  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      58.94
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2636.8236               Pseudo R2         =     0.0404

                                             (Std. Err. adjusted for 171 clusters in cowcode)
---------------------------------------------------------------------------------------------
                            |               Robust
              urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------------+----------------------------------------------------------------
  RESTRICT_HARASS_AMOUNT_bi |   .9151316   .1313672     6.97   0.000     .6576566    1.172607
RESTRICTsq_HARASS_AMOUNT_bi |  -.0818653   .0139594    -5.86   0.000    -.1092253   -.0545053
                      _cons |  -.1729655    .162694    -1.06   0.288    -.4918399    .1459088
----------------------------+----------------------------------------------------------------
                   /lnalpha |   1.274599   .1047944                      1.069206    1.479992
----------------------------+----------------------------------------------------------------
                      alpha |   3.577267   .3748775                      2.913066    4.392912
---------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2636.824       4    5281.647    5303.38
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2636.8236           4   5281.6473   5303.3796
.\Tables\Appendix_TableS4_HARASS_AMOUNT_bi.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12435.834  
Iteration 1:   log pseudolikelihood = -6396.4788  
Iteration 2:   log pseudolikelihood = -4293.9801  
Iteration 3:   log pseudolikelihood = -3082.3044  
Iteration 4:   log pseudolikelihood = -2938.8032  
Iteration 5:   log pseudolikelihood = -2936.8214  
Iteration 6:   log pseudolikelihood = -2936.8202  
Iteration 7:   log pseudolikelihood = -2936.8202  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2011.4427  
Iteration 1:   log pseudolikelihood = -1860.2458  
Iteration 2:   log pseudolikelihood = -1848.7808  
Iteration 3:   log pseudolikelihood = -1848.6912  
Iteration 4:   log pseudolikelihood = -1848.6912  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     275.39
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1848.6912               Pseudo R2         =     0.1358

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
   RESTRICT_HARASS_AMOUNT_bi |   .4879124   .0978595     4.99   0.000     .2961113    .6797134
 RESTRICTsq_HARASS_AMOUNT_bi |  -.0382119   .0105561    -3.62   0.000    -.0589015   -.0175223
                   PTS_Slag1 |   .6460984   .0981511     6.58   0.000     .4537257    .8384711
                hrgroupslag1 |   .0027186   .0029408     0.92   0.355    -.0030453    .0084825
                  hrnewslag1 |    .183433   .0555418     3.30   0.001      .074573     .292293
  protest_ClarkRegan_loglag1 |   .3391404   .0893428     3.80   0.000     .1640317    .5142491
      UCDP_armedConflictlag1 |   .3441389   .2708595     1.27   0.204    -.1867359    .8750138
         PR_freedomHouselag1 |    .645329    .242405     2.66   0.008     .1702239    1.120434
      PR_freedomHouselag1_sq |  -.0664905   .0289738    -2.29   0.022     -.123278    -.009703
   gdp_pc_constantUS2010lag1 |   .7740569   .3499558     2.21   0.027     .0881563    1.459958
gdp_pc_constantUS2010lag1_sq |  -.2663736   .1303694    -2.04   0.041     -.521893   -.0108542
                   KOFGIlag1 |   .1097079   .0453009     2.42   0.015     .0209199     .198496
                KOFGIlag1_sq |  -.0009593   .0004357    -2.20   0.028    -.0018132   -.0001054
              populationlag1 |   .1441421   .1700161     0.85   0.397    -.1890833    .4773674
                       _cons |  -6.520286    1.25507    -5.20   0.000    -8.980178   -4.060395
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5388528   .1274651                      .2890258    .7886797
-----------------------------+----------------------------------------------------------------
                       alpha |   1.714039   .2184801                      1.335126    2.200489
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1848.691      16    3729.382   3811.477
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_HARASS_AMOUNT_bi.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1931.9921  
Iteration 2:   log pseudolikelihood = -1882.4461  (not concave)
Iteration 3:   log pseudolikelihood = -1802.1521  
Iteration 4:   log pseudolikelihood = -1776.3432  
Iteration 5:   log pseudolikelihood =  -1775.756  
Iteration 6:   log pseudolikelihood = -1775.7544  
Iteration 7:   log pseudolikelihood = -1775.7544  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     252.99
Log pseudolikelihood = -1775.754                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
   RESTRICT_HARASS_AMOUNT_bi |   .3779671   .0834507     4.53   0.000     .2144067    .5415276
 RESTRICTsq_HARASS_AMOUNT_bi |   -.031661   .0094956    -3.33   0.001     -.050272     -.01305
                   PTS_Slag1 |    .496797   .0907257     5.48   0.000      .318978     .674616
                hrgroupslag1 |   .0009737    .002587     0.38   0.707    -.0040968    .0060441
                  hrnewslag1 |   .1206127   .0347106     3.47   0.001     .0525812    .1886443
  protest_ClarkRegan_loglag1 |   .3268783   .0728925     4.48   0.000     .1840116    .4697451
      UCDP_armedConflictlag1 |   .2368249   .2509584     0.94   0.345    -.2550446    .7286943
         PR_freedomHouselag1 |   .4289669   .2930328     1.46   0.143    -.1453669    1.003301
      PR_freedomHouselag1_sq |  -.0489407   .0335958    -1.46   0.145    -.1147873     .016906
   gdp_pc_constantUS2010lag1 |   .7263849   .3472784     2.09   0.036     .0457317    1.407038
gdp_pc_constantUS2010lag1_sq |  -.2689147    .133416    -2.02   0.044    -.5304053   -.0074241
                   KOFGIlag1 |   .0916299   .0414076     2.21   0.027     .0104725    .1727873
                KOFGIlag1_sq |  -.0008286   .0003967    -2.09   0.037    -.0016062    -.000051
              populationlag1 |   .1211496   .1747761     0.69   0.488    -.2214052    .4637044
                       _cons |   -4.23989   1.185096    -3.58   0.000    -6.562636   -1.917144
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .5905378   .5239065     1.13   0.260    -.4363001    1.617376
                     fhworst |  -1.840728   1.059922    -1.74   0.082    -3.918138    .2366817
      UCDP_armedConflictlag1 |  -.8091581   .4306596    -1.88   0.060    -1.653235    .0349193
                deathpenalty |   .1601969   .3139423     0.51   0.610    -.4551187    .7755125
           urgentActionslag1 |  -1.242349   .2278681    -5.45   0.000    -1.688962   -.7957359
                       _cons |   .3270404    .259872     1.26   0.208    -.1822994    .8363803
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1067179    .149867    -0.71   0.476    -.4004517    .1870159
-----------------------------+----------------------------------------------------------------
                       alpha |   .8987792   .1346973                      .6700173    1.205647
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1775.754      22    3595.509   3708.389
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_HARASS_AMOUNT_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x..x........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
   RESTRICT_HARASS_AMOUNT_bi |   .5948413   .1477577     4.03   0.000     .3052415    .8844412
 RESTRICTsq_HARASS_AMOUNT_bi |  -.0515396   .0186247    -2.77   0.006    -.0880433   -.0150358
                   PTS_Slag1 |   .5458329   .1199623     4.55   0.000     .3107111    .7809548
                hrgroupslag1 |  -.0003245   .0030251    -0.11   0.915    -.0062536    .0056045
                  hrnewslag1 |   .1227441   .0277132     4.43   0.000     .0684273    .1770609
  protest_ClarkRegan_loglag1 |   .3167903   .0983775     3.22   0.001     .1239738    .5096068
      UCDP_armedConflictlag1 |   .2855787   .2624934     1.09   0.277    -.2288988    .8000562
         PR_freedomHouselag1 |   .5058336   .3369915     1.50   0.133    -.1546577    1.166325
      PR_freedomHouselag1_sq |  -.0599475   .0400168    -1.50   0.134    -.1383789    .0184839
   gdp_pc_constantUS2010lag1 |   1.022002   .5256795     1.94   0.052    -.0083108    2.052315
gdp_pc_constantUS2010lag1_sq |  -.4734472   .2813764    -1.68   0.092    -1.024935    .0780403
                   KOFGIlag1 |   .0934287   .0855491     1.09   0.275    -.0742445    .2611019
                KOFGIlag1_sq |  -.0009199   .0007847    -1.17   0.241    -.0024579    .0006181
              populationlag1 |   .0221112   .2471053     0.09   0.929    -.4622062    .5064286
                       _cons |  -4.637751   2.169081    -2.14   0.033    -8.889071   -.3864302
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_HARASS_AMOUNT_bi RESTRICTsq_HARASS_AMOUNT_bi
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_HARASS_AMOUNT_bil2 RESTRICTsq_HARASS_AMOUNT_bil2
               RESTRICT_HARASS_AMOUNT_bil3
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .058426 (p = 0.8090)

symmetric es_ic[1,1]
           c1
r1  .05842607
.\Tables\Appendix_TableS4_HARASS_AMOUNT_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x..x........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
   RESTRICT_HARASS_AMOUNT_bi |   .6382139   .1489256     4.29   0.000      .346325    .9301028
 RESTRICTsq_HARASS_AMOUNT_bi |  -.0560777   .0183541    -3.06   0.002    -.0920511   -.0201042
                   PTS_Slag1 |   .7024689   .1988441     3.53   0.000     .3127416    1.092196
         PR_freedomHouselag1 |   .3414617   .4025811     0.85   0.396    -.4475828    1.130506
      PR_freedomHouselag1_sq |  -.0457995   .0478875    -0.96   0.339    -.1396572    .0480582
  protest_ClarkRegan_loglag1 |    .391419   .1611436     2.43   0.015     .0755834    .7072546
   gdp_pc_constantUS2010lag1 |   1.043522   .5366139     1.94   0.052    -.0082221    2.095266
gdp_pc_constantUS2010lag1_sq |  -.4500928   .2887681    -1.56   0.119    -1.016068    .1158823
                   KOFGIlag1 |   .0898735   .0879518     1.02   0.307    -.0825089     .262256
                KOFGIlag1_sq |  -.0008441   .0008144    -1.04   0.300    -.0024403    .0007521
                hrgroupslag1 |  -.0028546   .0038452    -0.74   0.458    -.0103911    .0046819
                  hrnewslag1 |   .1068655   .0324463     3.29   0.001      .043272     .170459
              populationlag1 |    .036133   .2583928     0.14   0.889    -.4703075    .5425736
      UCDP_armedConflictlag1 |     .14843   .3019336     0.49   0.623     -.443349     .740209
                       _cons |  -4.721586   2.309793    -2.04   0.041    -9.248698   -.1944745
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_HARASS_AMOUNT_bi RESTRICTsq_HARASS_AMOUNT_bi PTS_Slag1
               PR_freedomHouselag1 PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_HARASS_AMOUNT_bil2 RESTRICTsq_HARASS_AMOUNT_bil2
               RESTRICT_HARASS_AMOUNT_bil3 PTS_Slag2 PR_freedomHouselag2
               PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .070995 (p = 0.7899)

symmetric es_ic[1,1]
           c1
r1  .07099462
.\Tables\Appendix_TableS4_HARASS_AMOUNT_bi.doc
dir : seeout
(2,404 missing values generated)
(2,579 missing values generated)
(2,579 missing values generated)
(2,579 missing values generated)
(2,579 missing values generated)
(2,579 missing values generated)
(2,579 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood =  -5229.294  
Iteration 1:   log pseudolikelihood = -5228.9204  
Iteration 2:   log pseudolikelihood = -5228.9203  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3235.0249  
Iteration 1:   log pseudolikelihood = -2752.8167  
Iteration 2:   log pseudolikelihood = -2746.3213  
Iteration 3:   log pseudolikelihood = -2746.3144  
Iteration 4:   log pseudolikelihood = -2746.3144  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2646.7855  
Iteration 1:   log pseudolikelihood = -2623.9568  
Iteration 2:   log pseudolikelihood =  -2607.357  
Iteration 3:   log pseudolikelihood = -2607.3288  
Iteration 4:   log pseudolikelihood = -2607.3288  

Negative binomial regression                    Number of obs     =      1,688
                                                Wald chi2(2)      =      87.53
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2607.3288               Pseudo R2         =     0.0506

                                      (Std. Err. adjusted for 171 clusters in cowcode)
--------------------------------------------------------------------------------------
                     |               Robust
       urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------------+----------------------------------------------------------------
  RESTRICT_ARREST_bi |   .9489412   .1131474     8.39   0.000     .7271763    1.170706
RESTRICTsq_ARREST_bi |  -.0830885   .0120651    -6.89   0.000    -.1067356   -.0594414
               _cons |  -.3909181   .1608344    -2.43   0.015    -.7061478   -.0756884
---------------------+----------------------------------------------------------------
            /lnalpha |   1.202408   .1059777                       .994696    1.410121
---------------------+----------------------------------------------------------------
               alpha |   3.328123   .3527068                      2.703902    4.096451
--------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,688 -2746.314  -2607.329       4    5222.658   5244.383
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1688  -2746.3144  -2607.3288           4   5222.6576   5244.3828
.\Tables\Appendix_TableS4_ARREST_bi.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12328.271  
Iteration 1:   log pseudolikelihood = -6361.3517  
Iteration 2:   log pseudolikelihood = -4239.8387  
Iteration 3:   log pseudolikelihood = -3045.0819  
Iteration 4:   log pseudolikelihood = -2912.4983  
Iteration 5:   log pseudolikelihood = -2910.7187  
Iteration 6:   log pseudolikelihood = -2910.7179  
Iteration 7:   log pseudolikelihood = -2910.7179  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2008.9165  
Iteration 1:   log pseudolikelihood = -1854.7496  
Iteration 2:   log pseudolikelihood = -1843.2421  
Iteration 3:   log pseudolikelihood =  -1843.149  
Iteration 4:   log pseudolikelihood = -1843.1489  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     290.75
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1843.1489               Pseudo R2         =     0.1384

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
          RESTRICT_ARREST_bi |   .4990958   .1043216     4.78   0.000     .2946293    .7035624
        RESTRICTsq_ARREST_bi |  -.0380293   .0110543    -3.44   0.001    -.0596954   -.0163632
                   PTS_Slag1 |   .5974838   .0961989     6.21   0.000     .4089375    .7860301
                hrgroupslag1 |   .0027707   .0028146     0.98   0.325    -.0027459    .0082873
                  hrnewslag1 |   .1833303   .0525036     3.49   0.000     .0804251    .2862355
  protest_ClarkRegan_loglag1 |   .3279894   .0927924     3.53   0.000     .1461197    .5098591
      UCDP_armedConflictlag1 |   .4117483   .2734395     1.51   0.132    -.1241833    .9476799
         PR_freedomHouselag1 |   .6224442   .2389342     2.61   0.009     .1541418    1.090747
      PR_freedomHouselag1_sq |  -.0646748   .0287312    -2.25   0.024    -.1209869   -.0083627
   gdp_pc_constantUS2010lag1 |   .7815516   .3479813     2.25   0.025     .0995209    1.463582
gdp_pc_constantUS2010lag1_sq |  -.2677125   .1298423    -2.06   0.039    -.5221987   -.0132263
                   KOFGIlag1 |    .100752   .0451377     2.23   0.026     .0122839    .1892202
                KOFGIlag1_sq |  -.0009071   .0004288    -2.12   0.034    -.0017476   -.0000666
              populationlag1 |   .1059011   .1646803     0.64   0.520    -.2168662    .4286685
                       _cons |  -6.121397   1.258576    -4.86   0.000    -8.588161   -3.654634
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5206878   .1281444                      .2695295    .7718461
-----------------------------+----------------------------------------------------------------
                       alpha |   1.683185   .2156907                      1.309348    2.163757
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1843.149      16    3718.298   3800.392
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_ARREST_bi.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1928.7568  
Iteration 2:   log pseudolikelihood = -1882.0266  (not concave)
Iteration 3:   log pseudolikelihood = -1804.6435  
Iteration 4:   log pseudolikelihood = -1793.7873  (not concave)
Iteration 5:   log pseudolikelihood = -1788.4153  
Iteration 6:   log pseudolikelihood = -1773.0426  
Iteration 7:   log pseudolikelihood = -1772.9105  
Iteration 8:   log pseudolikelihood = -1772.9102  
Iteration 9:   log pseudolikelihood = -1772.9102  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     255.93
Log pseudolikelihood =  -1772.91                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
          RESTRICT_ARREST_bi |   .3853287   .0936325     4.12   0.000     .2018124    .5688451
        RESTRICTsq_ARREST_bi |   -.031248   .0102995    -3.03   0.002    -.0514345   -.0110614
                   PTS_Slag1 |   .4647296   .0890242     5.22   0.000     .2902454    .6392138
                hrgroupslag1 |   .0010817   .0024898     0.43   0.664    -.0037983    .0059617
                  hrnewslag1 |   .1234319   .0335907     3.67   0.000     .0575954    .1892684
  protest_ClarkRegan_loglag1 |   .3156519   .0746564     4.23   0.000     .1693281    .4619757
      UCDP_armedConflictlag1 |    .290422   .2534108     1.15   0.252    -.2062541    .7870981
         PR_freedomHouselag1 |   .4288065   .2922726     1.47   0.142    -.1440373     1.00165
      PR_freedomHouselag1_sq |  -.0490617   .0336484    -1.46   0.145    -.1150113     .016888
   gdp_pc_constantUS2010lag1 |   .7217131   .3468267     2.08   0.037     .0419452    1.401481
gdp_pc_constantUS2010lag1_sq |  -.2676148   .1330363    -2.01   0.044    -.5283612   -.0068684
                   KOFGIlag1 |   .0847229   .0411812     2.06   0.040     .0040092    .1654367
                KOFGIlag1_sq |  -.0007839   .0003898    -2.01   0.044    -.0015478   -.0000199
              populationlag1 |   .0863118   .1714333     0.50   0.615    -.2496913    .4223148
                       _cons |  -4.014691   1.179367    -3.40   0.001    -6.326208   -1.703174
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .5783112   .5230404     1.11   0.269    -.4468291    1.603452
                     fhworst |  -1.898706   1.125648    -1.69   0.092    -4.104935    .3075234
      UCDP_armedConflictlag1 |   -.750687   .4255423    -1.76   0.078    -1.584735    .0833607
                deathpenalty |   .1535073   .3126536     0.49   0.623    -.4592824    .7662971
           urgentActionslag1 |  -1.224066    .226417    -5.41   0.000    -1.667835   -.7802969
                       _cons |   .3036438   .2604456     1.17   0.244    -.2068203    .8141078
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1128784   .1539261    -0.73   0.463     -.414568    .1888112
-----------------------------+----------------------------------------------------------------
                       alpha |   .8932593   .1374959                      .6606256    1.207813
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85   -1772.91      22     3589.82     3702.7
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_ARREST_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,245
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
          RESTRICT_ARREST_bi |    .576408   .1400518     4.12   0.000     .3019116    .8509044
        RESTRICTsq_ARREST_bi |  -.0469662   .0161628    -2.91   0.004    -.0786447   -.0152878
                   PTS_Slag1 |   .4959802   .1238083     4.01   0.000     .2533203    .7386401
                hrgroupslag1 |   .0001988    .002891     0.07   0.945    -.0054675    .0058651
                  hrnewslag1 |   .1244029   .0269152     4.62   0.000     .0716501    .1771558
  protest_ClarkRegan_loglag1 |   .2930357   .0994393     2.95   0.003     .0981382    .4879331
      UCDP_armedConflictlag1 |   .3562976   .2453673     1.45   0.146    -.1246135    .8372088
         PR_freedomHouselag1 |   .4888809   .3404974     1.44   0.151    -.1784818    1.156244
      PR_freedomHouselag1_sq |  -.0584768   .0399543    -1.46   0.143    -.1367858    .0198322
   gdp_pc_constantUS2010lag1 |   .9760865   .5054964     1.93   0.053    -.0146682    1.966841
gdp_pc_constantUS2010lag1_sq |  -.4400203   .2617677    -1.68   0.093    -.9530755    .0730349
                   KOFGIlag1 |   .0897022    .081043     1.11   0.268    -.0691391    .2485436
                KOFGIlag1_sq |  -.0008899   .0007387    -1.20   0.228    -.0023377     .000558
              populationlag1 |   .0273243   .2533356     0.11   0.914    -.4692042    .5238529
                       _cons |  -4.514014   2.103546    -2.15   0.032    -8.636888   -.3911401
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_ARREST_bi RESTRICTsq_ARREST_bi
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_ARREST_bil2 RESTRICTsq_ARREST_bil2
               RESTRICT_ARREST_bil3

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .004553 (p = 0.9462)

symmetric es_ic[1,1]
           c1
r1  .00455253
.\Tables\Appendix_TableS4_ARREST_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,245
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
          RESTRICT_ARREST_bi |   .6009541   .1367238     4.40   0.000     .3329803    .8689278
        RESTRICTsq_ARREST_bi |  -.0491028    .015815    -3.10   0.002    -.0800996    -.018106
                   PTS_Slag1 |   .6510421   .1905134     3.42   0.001     .2776427    1.024441
         PR_freedomHouselag1 |   .3626022   .4126823     0.88   0.380    -.4462402    1.171445
      PR_freedomHouselag1_sq |  -.0487158   .0490199    -0.99   0.320    -.1447931    .0473615
  protest_ClarkRegan_loglag1 |   .3766835   .1708065     2.21   0.027     .0419089    .7114581
   gdp_pc_constantUS2010lag1 |   .9824489   .5153865     1.91   0.057      -.02769    1.992588
gdp_pc_constantUS2010lag1_sq |  -.4089244   .2673621    -1.53   0.126    -.9329445    .1150956
                   KOFGIlag1 |   .0838343   .0833365     1.01   0.314    -.0795022    .2471708
                KOFGIlag1_sq |  -.0007895   .0007686    -1.03   0.304    -.0022959    .0007169
                hrgroupslag1 |    -.00235   .0036936    -0.64   0.525    -.0095893    .0048893
                  hrnewslag1 |   .1071984   .0322751     3.32   0.001     .0439403    .1704564
              populationlag1 |   .0388821   .2591563     0.15   0.881     -.469055    .5468191
      UCDP_armedConflictlag1 |   .2326181   .2840112     0.82   0.413    -.3240336    .7892698
                       _cons |  -4.617947   2.233105    -2.07   0.039    -8.994752   -.2411412
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_ARREST_bi RESTRICTsq_ARREST_bi PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_ARREST_bil2 RESTRICTsq_ARREST_bil2 RESTRICT_ARREST_bil3 PTS_Slag2
               PR_freedomHouselag2 PR_freedomHouselag2_sq protest_ClarkRegan_loglag2

  Test of overidentifying restriction:

  Hansen's J chi2(1) =  .00103 (p = 0.9744)

symmetric es_ic[1,1]
           c1
r1  .00103038
.\Tables\Appendix_TableS4_ARREST_bi.doc
dir : seeout
(2,400 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5305.7324  
Iteration 1:   log pseudolikelihood = -5305.3977  
Iteration 2:   log pseudolikelihood = -5305.3976  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2649.6015  
Iteration 1:   log pseudolikelihood = -2628.5977  
Iteration 2:   log pseudolikelihood = -2608.9133  
Iteration 3:   log pseudolikelihood =  -2608.876  
Iteration 4:   log pseudolikelihood =  -2608.876  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      68.77
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood =  -2608.876               Pseudo R2         =     0.0506

                                       (Std. Err. adjusted for 171 clusters in cowcode)
---------------------------------------------------------------------------------------
                      |               Robust
        urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
  RESTRICT_SURVEIL_bi |   .9631614   .1328975     7.25   0.000      .702687    1.223636
RESTRICTsq_SURVEIL_bi |  -.0836625    .013954    -6.00   0.000    -.1110117   -.0563132
                _cons |  -.4137399   .1582446    -2.61   0.009    -.7238936   -.1035862
----------------------+----------------------------------------------------------------
             /lnalpha |   1.208766   .0994063                      1.013933    1.403599
----------------------+----------------------------------------------------------------
                alpha |   3.349349   .3329465                      2.756421     4.06982
---------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2608.876       4    5225.752   5247.484
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088   -2608.876           4   5225.7521   5247.4844
.\Tables\Appendix_TableS4_SURVEIL_bi.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12251.364  
Iteration 1:   log pseudolikelihood = -6413.5545  
Iteration 2:   log pseudolikelihood = -4245.5465  
Iteration 3:   log pseudolikelihood = -3027.3635  
Iteration 4:   log pseudolikelihood =  -2901.236  
Iteration 5:   log pseudolikelihood = -2899.4911  
Iteration 6:   log pseudolikelihood = -2899.4899  
Iteration 7:   log pseudolikelihood = -2899.4899  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2009.4111  
Iteration 1:   log pseudolikelihood = -1853.3498  
Iteration 2:   log pseudolikelihood = -1840.0929  
Iteration 3:   log pseudolikelihood = -1839.8854  
Iteration 4:   log pseudolikelihood = -1839.8853  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     291.74
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1839.8853               Pseudo R2         =     0.1399

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
         RESTRICT_SURVEIL_bi |   .5273457    .108049     4.88   0.000     .3155735    .7391179
       RESTRICTsq_SURVEIL_bi |  -.0407095   .0112674    -3.61   0.000    -.0627932   -.0186259
                   PTS_Slag1 |   .6146456   .0995262     6.18   0.000     .4195778    .8097134
                hrgroupslag1 |   .0015364   .0026743     0.57   0.566    -.0037052    .0067779
                  hrnewslag1 |    .192066   .0539839     3.56   0.000     .0862595    .2978726
  protest_ClarkRegan_loglag1 |   .3393904   .0931031     3.65   0.000     .1569117    .5218691
      UCDP_armedConflictlag1 |   .3326926   .2486928     1.34   0.181    -.1547363    .8201214
         PR_freedomHouselag1 |    .576146   .2292938     2.51   0.012     .1267385    1.025554
      PR_freedomHouselag1_sq |  -.0602664     .02742    -2.20   0.028    -.1140085   -.0065242
   gdp_pc_constantUS2010lag1 |   .7405387   .3290342     2.25   0.024     .0956435    1.385434
gdp_pc_constantUS2010lag1_sq |   -.254442   .1227951    -2.07   0.038     -.495116    -.013768
                   KOFGIlag1 |   .1047489   .0437286     2.40   0.017     .0190424    .1904555
                KOFGIlag1_sq |  -.0008999   .0004111    -2.19   0.029    -.0017057   -.0000941
              populationlag1 |   .1206507   .1654301     0.73   0.466    -.2035863    .4448877
                       _cons |  -6.282644   1.253497    -5.01   0.000    -8.739453   -3.825835
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5067307    .118871                      .2737478    .7397137
-----------------------------+----------------------------------------------------------------
                       alpha |   1.659856   .1973088                      1.314883    2.095335
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1839.885      16    3711.771   3793.865
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_SURVEIL_bi.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood =  -1817.012  
Iteration 2:   log pseudolikelihood = -1774.4595  
Iteration 3:   log pseudolikelihood = -1769.8085  
Iteration 4:   log pseudolikelihood = -1769.7505  
Iteration 5:   log pseudolikelihood = -1769.7505  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     239.24
Log pseudolikelihood =  -1769.75                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
         RESTRICT_SURVEIL_bi |    .415656   .0908711     4.57   0.000     .2375519      .59376
       RESTRICTsq_SURVEIL_bi |  -.0340096   .0098873    -3.44   0.001    -.0533884   -.0146308
                   PTS_Slag1 |   .4834529     .09186     5.26   0.000     .3034106    .6634952
                hrgroupslag1 |  -.0000901   .0023501    -0.04   0.969    -.0046962     .004516
                  hrnewslag1 |   .1287698   .0349518     3.68   0.000     .0602655    .1972742
  protest_ClarkRegan_loglag1 |    .329444   .0759086     4.34   0.000     .1806658    .4782222
      UCDP_armedConflictlag1 |   .2358245   .2362446     1.00   0.318    -.2272064    .6988554
         PR_freedomHouselag1 |   .3797315   .2777446     1.37   0.172    -.1646378    .9241009
      PR_freedomHouselag1_sq |   -.044509   .0318255    -1.40   0.162    -.1068858    .0178678
   gdp_pc_constantUS2010lag1 |   .6865967   .3335054     2.06   0.040     .0329382    1.340255
gdp_pc_constantUS2010lag1_sq |  -.2562159    .128165    -2.00   0.046    -.5074147   -.0050171
                   KOFGIlag1 |   .0887043   .0406809     2.18   0.029     .0089711    .1684375
                KOFGIlag1_sq |  -.0007797   .0003817    -2.04   0.041    -.0015279   -.0000316
              populationlag1 |   .0982015   .1728632     0.57   0.570    -.2406042    .4370072
                       _cons |  -4.174072   1.197232    -3.49   0.000    -6.520603   -1.827541
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .5815092   .5192551     1.12   0.263    -.4362121     1.59923
                     fhworst |  -1.919336   1.145883    -1.67   0.094    -4.165225    .3265539
      UCDP_armedConflictlag1 |  -.7773377   .4270435    -1.82   0.069    -1.614328    .0596521
                deathpenalty |   .1388945   .3143892     0.44   0.659    -.4772971    .7550861
           urgentActionslag1 |  -1.219483   .2273034    -5.37   0.000    -1.664989   -.7739762
                       _cons |   .3044423   .2637058     1.15   0.248    -.2124115    .8212961
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1249141   .1411177    -0.89   0.376    -.4014998    .1516716
-----------------------------+----------------------------------------------------------------
                       alpha |   .8825727   .1245467                      .6693155    1.163778
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85   -1769.75      22    3583.501   3696.381
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_SURVEIL_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
         RESTRICT_SURVEIL_bi |   .6544996   .1686849     3.88   0.000     .3238832     .985116
       RESTRICTsq_SURVEIL_bi |  -.0552866   .0192755    -2.87   0.004    -.0930658   -.0175073
                   PTS_Slag1 |   .5309472   .1364594     3.89   0.000     .2634916    .7984028
                hrgroupslag1 |  -.0012918   .0030876    -0.42   0.676    -.0073435    .0047599
                  hrnewslag1 |   .1323853   .0286497     4.62   0.000     .0762329    .1885377
  protest_ClarkRegan_loglag1 |    .299109   .1011366     2.96   0.003     .1008849    .4973332
      UCDP_armedConflictlag1 |   .2771988    .244058     1.14   0.256     -.201146    .7555436
         PR_freedomHouselag1 |   .4841745   .3526224     1.37   0.170    -.2069527    1.175302
      PR_freedomHouselag1_sq |  -.0600069   .0413689    -1.45   0.147    -.1410885    .0210746
   gdp_pc_constantUS2010lag1 |   .9978413   .4929018     2.02   0.043     .0317716    1.963911
gdp_pc_constantUS2010lag1_sq |  -.4552834   .2553611    -1.78   0.075     -.955782    .0452152
                   KOFGIlag1 |   .0937083   .0815529     1.15   0.251    -.0661323     .253549
                KOFGIlag1_sq |  -.0009069    .000739    -1.23   0.220    -.0023554    .0005416
              populationlag1 |  -.0123792   .2466298    -0.05   0.960    -.4957647    .4710062
                       _cons |  -4.705216   2.138226    -2.20   0.028    -8.896062   -.5143705
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_SURVEIL_bi RESTRICTsq_SURVEIL_bi
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_SURVEIL_bil2 RESTRICTsq_SURVEIL_bil2
               RESTRICT_SURVEIL_bil3

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .086399 (p = 0.7688)

symmetric es_ic[1,1]
           c1
r1  .08639901
.\Tables\Appendix_TableS4_SURVEIL_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.x.....x..........................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
         RESTRICT_SURVEIL_bi |    .703027   .1790785     3.93   0.000     .3520396    1.054014
       RESTRICTsq_SURVEIL_bi |  -.0601869   .0198235    -3.04   0.002    -.0990402   -.0213335
                   PTS_Slag1 |   .7043402   .2267636     3.11   0.002     .2598918    1.148789
         PR_freedomHouselag1 |   .3093791   .4410411     0.70   0.483    -.5550457    1.173804
      PR_freedomHouselag1_sq |  -.0452869   .0523099    -0.87   0.387    -.1478123    .0572386
  protest_ClarkRegan_loglag1 |   .3850193   .1664413     2.31   0.021     .0588004    .7112382
   gdp_pc_constantUS2010lag1 |     1.0161   .4960674     2.05   0.041      .043826    1.988375
gdp_pc_constantUS2010lag1_sq |   -.425209   .2581902    -1.65   0.100    -.9312525    .0808345
                   KOFGIlag1 |   .0892504   .0851327     1.05   0.294    -.0776066    .2561074
                KOFGIlag1_sq |   -.000816   .0007781    -1.05   0.294    -.0023411    .0007092
                hrgroupslag1 |  -.0042018   .0042677    -0.98   0.325    -.0125664    .0041627
                  hrnewslag1 |   .1150727    .033959     3.39   0.001     .0485143    .1816311
              populationlag1 |  -.0017971   .2223526    -0.01   0.994    -.4376001    .4340059
      UCDP_armedConflictlag1 |   .1294094    .291353     0.44   0.657    -.4416319    .7004508
                       _cons |  -4.814183    2.33509    -2.06   0.039    -9.390876   -.2374897
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_SURVEIL_bi RESTRICTsq_SURVEIL_bi PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_SURVEIL_bil2 RESTRICTsq_SURVEIL_bil2 RESTRICT_SURVEIL_bil3 PTS_Slag2
               PR_freedomHouselag2 PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .143058 (p = 0.7053)

symmetric es_ic[1,1]
           c1
r1  .14305791
.\Tables\Appendix_TableS4_SURVEIL_bi.doc
dir : seeout
(2,400 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
(2,575 missing values generated)
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5279.4203  
Iteration 1:   log pseudolikelihood = -5279.1349  
Iteration 2:   log pseudolikelihood = -5279.1349  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2647.4039  
Iteration 1:   log pseudolikelihood = -2626.1913  
Iteration 2:   log pseudolikelihood = -2605.3172  
Iteration 3:   log pseudolikelihood = -2605.2743  
Iteration 4:   log pseudolikelihood = -2605.2743  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      72.11
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2605.2743               Pseudo R2         =     0.0519

                                       (Std. Err. adjusted for 171 clusters in cowcode)
---------------------------------------------------------------------------------------
                      |               Robust
        urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
  RESTRICT_KILLING_bi |   .9486008   .1299582     7.30   0.000     .6938875    1.203314
RESTRICTsq_KILLING_bi |  -.0813423   .0134194    -6.06   0.000    -.1076437   -.0550408
                _cons |   -.441204   .1603257    -2.75   0.006    -.7554366   -.1269714
----------------------+----------------------------------------------------------------
             /lnalpha |   1.198927   .1029858                      .9970787    1.400776
----------------------+----------------------------------------------------------------
                alpha |   3.316557   .3415585                      2.710353    4.058347
---------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2605.274       4    5218.549   5240.281
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2605.2743           4   5218.5487    5240.281
.\Tables\Appendix_TableS4_KILLING_bi.doc
dir : seeout

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12231.371  
Iteration 1:   log pseudolikelihood = -6457.2852  
Iteration 2:   log pseudolikelihood = -4253.4154  
Iteration 3:   log pseudolikelihood = -3027.6617  
Iteration 4:   log pseudolikelihood = -2909.0797  
Iteration 5:   log pseudolikelihood = -2907.2336  
Iteration 6:   log pseudolikelihood =  -2907.232  
Iteration 7:   log pseudolikelihood =  -2907.232  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2009.4393  
Iteration 1:   log pseudolikelihood = -1853.2753  
Iteration 2:   log pseudolikelihood = -1840.3508  
Iteration 3:   log pseudolikelihood = -1840.1593  
Iteration 4:   log pseudolikelihood = -1840.1592  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     297.83
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1840.1592               Pseudo R2         =     0.1398

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
         RESTRICT_KILLING_bi |   .5143341   .1094653     4.70   0.000      .299786    .7288822
       RESTRICTsq_KILLING_bi |  -.0392041   .0111357    -3.52   0.000    -.0610297   -.0173785
                   PTS_Slag1 |   .5997756   .0974031     6.16   0.000      .408869    .7906823
                hrgroupslag1 |   .0023452   .0026637     0.88   0.379    -.0028754    .0075659
                  hrnewslag1 |   .1855062   .0518794     3.58   0.000     .0838245    .2871879
  protest_ClarkRegan_loglag1 |   .3256251   .0924641     3.52   0.000     .1443988    .5068514
      UCDP_armedConflictlag1 |   .3505416   .2489718     1.41   0.159    -.1374342    .8385174
         PR_freedomHouselag1 |   .5716779   .2294751     2.49   0.013      .121915    1.021441
      PR_freedomHouselag1_sq |   -.058677   .0276231    -2.12   0.034    -.1128172   -.0045368
   gdp_pc_constantUS2010lag1 |   .7446891   .3267509     2.28   0.023      .104269    1.385109
gdp_pc_constantUS2010lag1_sq |  -.2576451   .1232444    -2.09   0.037    -.4991998   -.0160904
                   KOFGIlag1 |   .1020936   .0432314     2.36   0.018     .0173617    .1868255
                KOFGIlag1_sq |  -.0009012   .0004085    -2.21   0.027     -.001702   -.0001005
              populationlag1 |   .0979293   .1672532     0.59   0.558     -.229881    .4257396
                       _cons |  -6.152772   1.234392    -4.98   0.000    -8.572137   -3.733407
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5092883   .1192057                      .2756494    .7429271
-----------------------------+----------------------------------------------------------------
                       alpha |   1.664106    .198371                      1.317386     2.10208
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1840.159      16    3712.318   3794.413
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_KILLING_bi.doc
dir : seeout

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1816.9973  
Iteration 2:   log pseudolikelihood = -1774.1958  
Iteration 3:   log pseudolikelihood = -1769.4428  
Iteration 4:   log pseudolikelihood = -1769.3916  
Iteration 5:   log pseudolikelihood = -1769.3916  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     247.52
Log pseudolikelihood = -1769.392                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
         RESTRICT_KILLING_bi |   .4138951   .0926965     4.47   0.000     .2322134    .5955769
       RESTRICTsq_KILLING_bi |    -.03364    .009739    -3.45   0.001    -.0527282   -.0145519
                   PTS_Slag1 |   .4733828   .0905638     5.23   0.000     .2958809    .6508846
                hrgroupslag1 |   .0007081   .0023598     0.30   0.764     -.003917    .0053332
                  hrnewslag1 |   .1239162   .0339516     3.65   0.000     .0573723    .1904601
  protest_ClarkRegan_loglag1 |   .3154563   .0751353     4.20   0.000     .1681938    .4627189
      UCDP_armedConflictlag1 |   .2340357   .2358646     0.99   0.321    -.2282503    .6963218
         PR_freedomHouselag1 |   .3708084   .2740918     1.35   0.176    -.1664016    .9080184
      PR_freedomHouselag1_sq |  -.0424089   .0315862    -1.34   0.179    -.1043166    .0194989
   gdp_pc_constantUS2010lag1 |    .686503   .3296571     2.08   0.037      .040387    1.332619
gdp_pc_constantUS2010lag1_sq |  -.2578794   .1275482    -2.02   0.043    -.5078692   -.0078897
                   KOFGIlag1 |    .085677    .039937     2.15   0.032      .007402    .1639521
                KOFGIlag1_sq |   -.000778   .0003772    -2.06   0.039    -.0015174   -.0000387
              populationlag1 |   .0688748   .1752983     0.39   0.694    -.2747036    .4124532
                       _cons |  -4.038824    1.17248    -3.44   0.001    -6.336842   -1.740806
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .5887241   .5175222     1.14   0.255    -.4256007    1.603049
                     fhworst |  -1.818775   1.053447    -1.73   0.084    -3.883493    .2459436
      UCDP_armedConflictlag1 |  -.7964668    .428918    -1.86   0.063    -1.637131    .0441971
                deathpenalty |   .1391702   .3137538     0.44   0.657    -.4757759    .7541163
           urgentActionslag1 |  -1.218563   .2287249    -5.33   0.000    -1.666855   -.7702701
                       _cons |   .3045402   .2619703     1.16   0.245    -.2089122    .8179925
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1268695   .1407671    -0.90   0.367    -.4027679    .1490289
-----------------------------+----------------------------------------------------------------
                       alpha |   .8808486   .1239945                      .6684672    1.160707
----------------------------------------------------------------------------------------------

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1769.392      22    3582.783   3695.663
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.
.\Tables\Appendix_TableS4_KILLING_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
         RESTRICT_KILLING_bi |   .6242198   .1600623     3.90   0.000     .3105034    .9379362
       RESTRICTsq_KILLING_bi |  -.0512164    .017108    -2.99   0.003    -.0847474   -.0176854
                   PTS_Slag1 |   .5232646   .1325846     3.95   0.000     .2634034    .7831257
                hrgroupslag1 |  -.0002186   .0028089    -0.08   0.938     -.005724    .0052867
                  hrnewslag1 |   .1317036   .0297642     4.42   0.000     .0733667    .1900404
  protest_ClarkRegan_loglag1 |    .283165   .1033971     2.74   0.006     .0805105    .4858195
      UCDP_armedConflictlag1 |   .2861402   .2432844     1.18   0.240    -.1906884    .7629689
         PR_freedomHouselag1 |   .4544293   .3552748     1.28   0.201    -.2418965    1.150755
      PR_freedomHouselag1_sq |  -.0567522   .0419064    -1.35   0.176    -.1388873    .0253829
   gdp_pc_constantUS2010lag1 |   1.040995   .5132455     2.03   0.043     .0350527    2.046938
gdp_pc_constantUS2010lag1_sq |  -.4666516   .2621538    -1.78   0.075    -.9804636    .0471603
                   KOFGIlag1 |   .0864244   .0816397     1.06   0.290    -.0735865    .2464352
                KOFGIlag1_sq |  -.0008896   .0007475    -1.19   0.234    -.0023547    .0005756
              populationlag1 |   -.027265   .2455752    -0.11   0.912    -.5085836    .4540536
                       _cons |  -4.332157   2.132088    -2.03   0.042    -8.510973   -.1533404
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_KILLING_bi RESTRICTsq_KILLING_bi
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_KILLING_bil2 RESTRICTsq_KILLING_bil2
               RESTRICT_KILLING_bil3
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .078721 (p = 0.7790)

symmetric es_ic[1,1]
           c1
r1  .07872134
.\Tables\Appendix_TableS4_KILLING_bi.doc
dir : seeout
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x...........................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
         RESTRICT_KILLING_bi |   .6695398   .1761073     3.80   0.000     .3243759    1.014704
       RESTRICTsq_KILLING_bi |  -.0552057   .0172023    -3.21   0.001    -.0889215   -.0214899
                   PTS_Slag1 |   .7112296   .2355383     3.02   0.003      .249583    1.172876
         PR_freedomHouselag1 |   .2718714   .4458657     0.61   0.542    -.6020094    1.145752
      PR_freedomHouselag1_sq |  -.0421633    .052548    -0.80   0.422    -.1451555    .0608289
  protest_ClarkRegan_loglag1 |   .3507515    .174851     2.01   0.045     .0080498    .6934531
   gdp_pc_constantUS2010lag1 |   1.053747   .5271485     2.00   0.046     .0205547    2.086939
gdp_pc_constantUS2010lag1_sq |  -.4344748   .2687224    -1.62   0.106    -.9611609    .0922114
                   KOFGIlag1 |   .0804302   .0832585     0.97   0.334    -.0827533    .2436138
                KOFGIlag1_sq |  -.0007861   .0007691    -1.02   0.307    -.0022935    .0007213
                hrgroupslag1 |  -.0031483   .0039765    -0.79   0.429    -.0109421    .0046454
                  hrnewslag1 |   .1149716    .034493     3.33   0.001     .0473665    .1825766
              populationlag1 |  -.0220955   .2572021    -0.09   0.932    -.5262023    .4820113
      UCDP_armedConflictlag1 |   .1310175   .3031776     0.43   0.666    -.4631996    .7252346
                       _cons |  -4.390434   2.283099    -1.92   0.054    -8.865226    .0843578
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_KILLING_bi RESTRICTsq_KILLING_bi PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1
               RESTRICT_KILLING_bil2 RESTRICTsq_KILLING_bil2 RESTRICT_KILLING_bil3 PTS_Slag2
               PR_freedomHouselag2 PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .099375 (p = 0.7526)

symmetric es_ic[1,1]
           c1
r1  .09937518
.\Tables\Appendix_TableS4_KILLING_bi.doc
dir : seeout

. 
. 
. 
. *** Leave out packages of restrictions
. 
. capture drop RESTRICT_1

. capture drop RESTRICT_2

. capture drop RESTRICT_3

. capture drop RESTRICTsq_1

. capture drop RESTRICTsq_2

. capture drop RESTRICTsq_3

. 
. gen RESTRICT_1 = RESTRICT_COUNTdomlag1 - VISIT_RESTRICTlag1 - TRAVEL_RESTRICTlag1 -  FUNDING_INTlag1 - FUNDING_DOMlag1
(2,576 missing values generated)

. gen RESTRICTsq_1 = RESTRICT_1 * RESTRICT_1
(2,576 missing values generated)

. label var RESTRICT_1 "Restrictions (-- visit, travel, funding)"

. label var RESTRICTsq_1 "Restrictions sq."

. gen RESTRICT_2 = RESTRICT_COUNTdomlag1 - REGISTRATIONlag1 - CENSORlag1  - SURVEIL_bilag1 - SOME_BANNEDlag1
(2,575 missing values generated)

. gen RESTRICTsq_2 = RESTRICT_2 * RESTRICT_2
(2,575 missing values generated)

. label var RESTRICT_2 "Restrictions (-- registr., censor, surveil., banned)"

. label var RESTRICTsq_2 "Restrictions sq."

. gen RESTRICT_3 = RESTRICT_COUNTdomlag1 - HARASS_AMOUNT_bilag1 - ARREST_bilag1 - KILLING_bilag1
(2,579 missing values generated)

. gen RESTRICTsq_3 = RESTRICT_3 * RESTRICT_3
(2,579 missing values generated)

. label var RESTRICT_3 "Restrictions (-- harass, arrest, killing)"

. label var RESTRICTsq_3 "Restrictions sq."

. 
. sort cowcode YEAR

. gen RESTRICT_1lag2 = RESTRICT_1[_n-1]
(2,576 missing values generated)

. gen RESTRICTsq_1lag2 = RESTRICTsq_1[_n-1]
(2,576 missing values generated)

. gen RESTRICT_1lag3 = RESTRICT_1[_n-2]
(2,576 missing values generated)

. gen RESTRICTsq_1lag3 = RESTRICTsq_1[_n-2]
(2,576 missing values generated)

. 
. gen RESTRICT_2lag2 = RESTRICT_2[_n-1]
(2,575 missing values generated)

. gen RESTRICTsq_2lag2 = RESTRICTsq_2[_n-1]
(2,575 missing values generated)

. gen RESTRICT_2lag3 = RESTRICT_2[_n-2]
(2,575 missing values generated)

. gen RESTRICTsq_2lag3 = RESTRICTsq_2[_n-2]
(2,575 missing values generated)

. 
. gen RESTRICT_3lag2 = RESTRICT_3[_n-1]
(2,579 missing values generated)

. gen RESTRICTsq_3lag2 = RESTRICTsq_3[_n-1]
(2,579 missing values generated)

. gen RESTRICT_3lag3 = RESTRICT_3[_n-2]
(2,579 missing values generated)

. gen RESTRICTsq_3lag3 = RESTRICTsq_3[_n-2]
(2,579 missing values generated)

. 
. 
. * Leave out: VISIT_RESTRICTlag1 - TRAVEL_RESTRICTlag1 -  FUNDING_INTlag1 - FUNDING_DOMlag1
. 
.         xtset cowcode YEAR
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

.         
.         * Negative binomial 
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_1 RESTRICTsq_1, vce(cluster cowcode) ;

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5235.6671  
Iteration 1:   log pseudolikelihood = -5235.2801  
Iteration 2:   log pseudolikelihood =   -5235.28  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3236.4265  
Iteration 1:   log pseudolikelihood = -2751.6549  
Iteration 2:   log pseudolikelihood = -2745.0628  
Iteration 3:   log pseudolikelihood = -2745.0556  
Iteration 4:   log pseudolikelihood = -2745.0556  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2647.7554  
Iteration 1:   log pseudolikelihood = -2624.4829  
Iteration 2:   log pseudolikelihood = -2609.6109  
Iteration 3:   log pseudolikelihood = -2609.5837  
Iteration 4:   log pseudolikelihood = -2609.5837  

Negative binomial regression                    Number of obs     =      1,690
                                                Wald chi2(2)      =      62.82
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2609.5837               Pseudo R2         =     0.0494

                               (Std. Err. adjusted for 171 clusters in cowcode)
-------------------------------------------------------------------------------
              |               Robust
urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
   RESTRICT_1 |   .9483864   .1333992     7.11   0.000     .6869288    1.209844
 RESTRICTsq_1 |  -.0826888   .0137118    -6.03   0.000    -.1095636   -.0558141
        _cons |  -.3112639   .1612497    -1.93   0.054    -.6273075    .0047797
--------------+----------------------------------------------------------------
     /lnalpha |   1.210506    .096713                      1.020952     1.40006
--------------+----------------------------------------------------------------
        alpha |   3.355182   .3244897                      2.775836    4.055443
-------------------------------------------------------------------------------

.         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,690 -2745.056  -2609.584       4    5227.167   5248.897
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         mat list es_ic

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1690  -2745.0556  -2609.5837           4   5227.1673   5248.8972

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4m.doc", replace ///
>         keep(RESTRICT_1 RESTRICTsq_1) ///
>         ctitle("Model 1") label eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4m.doc
dir : seeout

. 
.         * Negative binomial
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_1 RESTRICTsq_1
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, vce(cluster cowcode);

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -11933.579  
Iteration 1:   log pseudolikelihood =  -6483.466  
Iteration 2:   log pseudolikelihood = -4181.0936  
Iteration 3:   log pseudolikelihood = -2936.4067  
Iteration 4:   log pseudolikelihood = -2847.8803  
Iteration 5:   log pseudolikelihood = -2846.2557  
Iteration 6:   log pseudolikelihood = -2846.2545  
Iteration 7:   log pseudolikelihood = -2846.2545  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood =  -2007.035  
Iteration 1:   log pseudolikelihood = -1850.4826  
Iteration 2:   log pseudolikelihood = -1837.8005  
Iteration 3:   log pseudolikelihood = -1837.6467  
Iteration 4:   log pseudolikelihood = -1837.6467  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     261.87
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1837.6467               Pseudo R2         =     0.1410

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_1 |    .530597   .1080574     4.91   0.000     .3188083    .7423857
                RESTRICTsq_1 |  -.0408377   .0109184    -3.74   0.000    -.0622373    -.019438
                   PTS_Slag1 |   .6188104    .101749     6.08   0.000     .4193859    .8182348
                hrgroupslag1 |   .0018148   .0026799     0.68   0.498    -.0034378    .0070673
                  hrnewslag1 |   .1861604   .0540477     3.44   0.001     .0802289    .2920919
  protest_ClarkRegan_loglag1 |   .3207821   .0915552     3.50   0.000     .1413372    .5002271
      UCDP_armedConflictlag1 |   .4306452   .2407734     1.79   0.074     -.041262    .9025525
         PR_freedomHouselag1 |   .5985821   .2238086     2.67   0.007     .1599254    1.037239
      PR_freedomHouselag1_sq |  -.0621379   .0268339    -2.32   0.021    -.1147314   -.0095444
   gdp_pc_constantUS2010lag1 |   .7024907   .3318803     2.12   0.034     .0520173    1.352964
gdp_pc_constantUS2010lag1_sq |  -.2489266   .1208342    -2.06   0.039    -.4857573    -.012096
                   KOFGIlag1 |   .0853402   .0462983     1.84   0.065    -.0054029    .1760833
                KOFGIlag1_sq |  -.0007299   .0004354    -1.68   0.094    -.0015833    .0001234
              populationlag1 |   .0945333   .1668968     0.57   0.571    -.2325785    .4216451
                       _cons |  -5.799474   1.306124    -4.44   0.000    -8.359431   -3.239518
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .4851244   .1183593                      .2531444    .7171043
-----------------------------+----------------------------------------------------------------
                       alpha |   1.624377   .1922601                      1.288069    2.048493
----------------------------------------------------------------------------------------------

.         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1837.647      16    3707.293   3789.388
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4m.doc", append ///
>         keep(RESTRICT_1 RESTRICTsq_1) ///
>         ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4m.doc
dir : seeout

. 
.         * Zero inflated negative binomial 
.         #delimit ;
delimiter now ;
.         zinb urgentActions RESTRICT_1 RESTRICTsq_1
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, 
>         inflate(fhbest fhworst UCDP_armedConflictlag1 deathpenalty urgentActionslag1) 
>         vce(cluster cowcode);

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1813.7516  
Iteration 2:   log pseudolikelihood = -1770.8357  
Iteration 3:   log pseudolikelihood = -1766.3718  
Iteration 4:   log pseudolikelihood = -1766.2992  
Iteration 5:   log pseudolikelihood = -1766.2992  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     235.90
Log pseudolikelihood = -1766.299                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
                  RESTRICT_1 |   .4287416   .0909961     4.71   0.000     .2503925    .6070907
                RESTRICTsq_1 |  -.0351148   .0094735    -3.71   0.000    -.0536825   -.0165471
                   PTS_Slag1 |   .4815918    .094898     5.07   0.000     .2955951    .6675885
                hrgroupslag1 |   .0002889   .0024176     0.12   0.905    -.0044495    .0050273
                  hrnewslag1 |   .1255511   .0346438     3.62   0.000     .0576506    .1934517
  protest_ClarkRegan_loglag1 |   .3120177   .0750884     4.16   0.000     .1648471    .4591883
      UCDP_armedConflictlag1 |   .2971545   .2308164     1.29   0.198    -.1552374    .7495464
         PR_freedomHouselag1 |   .3686582    .267178     1.38   0.168     -.155001    .8923174
      PR_freedomHouselag1_sq |  -.0429011     .03084    -1.39   0.164    -.1033465    .0175443
   gdp_pc_constantUS2010lag1 |    .656514   .3290898     1.99   0.046     .0115098    1.301518
gdp_pc_constantUS2010lag1_sq |   -.251065   .1253763    -2.00   0.045     -.496798    -.005332
                   KOFGIlag1 |    .069761   .0443841     1.57   0.116    -.0172302    .1567522
                KOFGIlag1_sq |  -.0006096   .0004168    -1.46   0.144    -.0014265    .0002074
              populationlag1 |   .0833155   .1755059     0.47   0.635    -.2606697    .4273008
                       _cons |  -3.654274   1.270216    -2.88   0.004    -6.143852   -1.164697
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .6746664   .5015454     1.35   0.179    -.3083445    1.657677
                     fhworst |  -1.828787   1.139551    -1.60   0.109    -4.062265    .4046914
      UCDP_armedConflictlag1 |  -.8083926   .4447357    -1.82   0.069    -1.680059    .0632734
                deathpenalty |   .1583588   .3181926     0.50   0.619    -.4652872    .7820049
           urgentActionslag1 |  -1.225688   .2307781    -5.31   0.000    -1.678005    -.773371
                       _cons |   .2877276   .2706279     1.06   0.288    -.2426934    .8181486
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.1480689    .144809    -1.02   0.307    -.4318894    .1357516
-----------------------------+----------------------------------------------------------------
                       alpha |   .8623717   .1248792                      .6492812    1.145397
----------------------------------------------------------------------------------------------

.  // vuong ;
>         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1766.299      22    3576.598   3689.478
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4m.doc", append ///
>         keep(RESTRICT_1 RESTRICTsq_1) ///
>         ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4m.doc
dir : seeout

.         
.         * GMM
.         xtset, clear

.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1
>         ( RESTRICT_1 RESTRICTsq_1 =
>                 RESTRICT_1lag2 RESTRICT_1lag3 RESTRICTsq_1lag2)
>                 , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_1 |   .6773111   .1717933     3.94   0.000     .3406025     1.01402
                RESTRICTsq_1 |  -.0571743   .0171441    -3.33   0.001    -.0907761   -.0235726
                   PTS_Slag1 |   .4840154   .1242025     3.90   0.000     .2405829    .7274478
                hrgroupslag1 |  -.0005061   .0029625    -0.17   0.864    -.0063125    .0053003
                  hrnewslag1 |   .1176173   .0263258     4.47   0.000     .0660196    .1692149
  protest_ClarkRegan_loglag1 |   .3267175   .0929224     3.52   0.000      .144593    .5088419
      UCDP_armedConflictlag1 |   .3971675   .2387453     1.66   0.096    -.0707647    .8650997
         PR_freedomHouselag1 |   .4624764   .3420586     1.35   0.176    -.2079463    1.132899
      PR_freedomHouselag1_sq |  -.0551125   .0386215    -1.43   0.154    -.1308093    .0205842
   gdp_pc_constantUS2010lag1 |   .9303356   .4918325     1.89   0.059    -.0336383     1.89431
gdp_pc_constantUS2010lag1_sq |  -.4219668   .2549391    -1.66   0.098    -.9216382    .0777045
                   KOFGIlag1 |   .0779077   .0851284     0.92   0.360    -.0889409    .2447563
                KOFGIlag1_sq |  -.0007412   .0007731    -0.96   0.338    -.0022565     .000774
              populationlag1 |   .0494198   .2580622     0.19   0.848    -.4563729    .5552124
                       _cons |  -4.320982   2.134363    -2.02   0.043    -8.504256   -.1377072
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_1 RESTRICTsq_1
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_1lag2 RESTRICT_1lag3 RESTRICTsq_1lag2

.         #delimit cr
delimiter now cr
.         
.         estat overid

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .109996 (p = 0.7401)

.         mat es_ic = r(J) 

.         matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .10999602

.         local J: display %4.1f es_ic[1,1]

.         outreg2 using ".\Tables\Appendix_TableS4m.doc", append ///
>         keep( RESTRICT_1 RESTRICTsq_1) ///
>         ctitle("Model 4") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Hansen's J, `J')
.\Tables\Appendix_TableS4m.doc
dir : seeout

. 
.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>                 gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>                 KOFGIlag1 KOFGIlag1_sq 
>                 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1 
>                 populationlag1  UCDP_armedConflictlag1 
>                 (RESTRICT_1 RESTRICTsq_1 
>                 PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
>                 = RESTRICT_1lag2 RESTRICT_1lag3 RESTRICTsq_1lag2
>                 PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq)
>         , twostep vce(boot, reps(50) cl(cowcode) seed(2)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.................................x................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_1 |   .7049858   .2129226     3.31   0.001     .2876651    1.122307
                RESTRICTsq_1 |  -.0609637    .022479    -2.71   0.007    -.1050217   -.0169057
                   PTS_Slag1 |   .6261907   .2169569     2.89   0.004     .2009629    1.051418
         PR_freedomHouselag1 |   .2611511   .3733062     0.70   0.484    -.4705157    .9928179
      PR_freedomHouselag1_sq |  -.0360556   .0469028    -0.77   0.442    -.1279835    .0558723
   gdp_pc_constantUS2010lag1 |   .9442828   .5583251     1.69   0.091    -.1500143     2.03858
gdp_pc_constantUS2010lag1_sq |  -.4057143    .323799    -1.25   0.210    -1.040349    .2289201
                   KOFGIlag1 |   .0774878    .087445     0.89   0.376    -.0939013    .2488769
                KOFGIlag1_sq |  -.0007085   .0007844    -0.90   0.366    -.0022459    .0008288
                hrgroupslag1 |  -.0022512   .0038109    -0.59   0.555    -.0097203     .005218
                  hrnewslag1 |   .1091115    .038325     2.85   0.004     .0339959    .1842271
  protest_ClarkRegan_loglag1 |   .3162086   .1157088     2.73   0.006     .0894236    .5429935
              populationlag1 |   .0477517   .2477584     0.19   0.847    -.4378459    .5333493
      UCDP_armedConflictlag1 |   .2812811   .3179799     0.88   0.376     -.341948    .9045102
                       _cons |  -4.267759   2.388159    -1.79   0.074    -8.948465    .4129464
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_1 RESTRICTsq_1 PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1 populationlag1
               UCDP_armedConflictlag1 RESTRICT_1lag2 RESTRICT_1lag3 RESTRICTsq_1lag2
               PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

.         #delimit cr
delimiter now cr
. 
.         estat overid 

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .180409 (p = 0.6710)

.         mat es_ic = r(J) 

.          matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .18040888

.         local J: display %4.1f es_ic[1,1]

.         outreg2 using ".\Tables\Appendix_TableS4m.doc", append ///
>          ctitle("Model 5") label  eqdrop(lnalpha) dec(3) ///
>          keep(RESTRICT_1 RESTRICTsq_1) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Hansen's J, `J')
.\Tables\Appendix_TableS4m.doc
dir : seeout

.         
.         
.         
. 
.         
. 
. ** REGISTRATION_PROBLEMSlag1 - CENSORlag1  - SURVEIL_bilag1 -- SOME_BANNEDlag1
. 
. 
.         xtset cowcode YEAR
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

.         
.         * Negative binomial 
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_2 RESTRICTsq_2, vce(cluster cowcode) ;

Fitting Poisson model:

Iteration 0:   log pseudolikelihood =  -5350.645  
Iteration 1:   log pseudolikelihood =  -5350.614  
Iteration 2:   log pseudolikelihood =  -5350.614  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3238.3361  
Iteration 1:   log pseudolikelihood = -2754.4031  
Iteration 2:   log pseudolikelihood = -2747.8159  
Iteration 3:   log pseudolikelihood = -2747.8088  
Iteration 4:   log pseudolikelihood = -2747.8088  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2679.4193  
Iteration 1:   log pseudolikelihood = -2662.6307  
Iteration 2:   log pseudolikelihood = -2653.8451  
Iteration 3:   log pseudolikelihood = -2653.8255  
Iteration 4:   log pseudolikelihood = -2653.8255  

Negative binomial regression                    Number of obs     =      1,691
                                                Wald chi2(2)      =      35.90
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2653.8255               Pseudo R2         =     0.0342

                               (Std. Err. adjusted for 171 clusters in cowcode)
-------------------------------------------------------------------------------
              |               Robust
urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
   RESTRICT_2 |   .7369428   .1401677     5.26   0.000     .4622192    1.011666
 RESTRICTsq_2 |  -.0707464   .0172076    -4.11   0.000    -.1044726   -.0370202
        _cons |  -.0096652   .1848639    -0.05   0.958    -.3719918    .3526615
--------------+----------------------------------------------------------------
     /lnalpha |   1.300927   .1116698                      1.082058    1.519795
--------------+----------------------------------------------------------------
        alpha |   3.672698   .4101295                      2.950745     4.57129
-------------------------------------------------------------------------------

.         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,691 -2747.809  -2653.825       4    5315.651   5337.383
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         mat list es_ic

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1691  -2747.8088  -2653.8255           4   5315.6509   5337.3832

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4l.doc", replace ///
>         keep(RESTRICT_2 RESTRICTsq_2) ///
>         ctitle("Model 1") label eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4l.doc
dir : seeout

. 
. 
.         * Negative binomial
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_2 RESTRICTsq_2
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, vce(cluster cowcode);

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12147.312  
Iteration 1:   log pseudolikelihood = -6555.2937  
Iteration 2:   log pseudolikelihood = -4301.8296  
Iteration 3:   log pseudolikelihood = -3050.5406  
Iteration 4:   log pseudolikelihood = -2954.2807  
Iteration 5:   log pseudolikelihood = -2952.2097  
Iteration 6:   log pseudolikelihood = -2952.2078  
Iteration 7:   log pseudolikelihood = -2952.2078  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood =  -2011.707  
Iteration 1:   log pseudolikelihood = -1870.8815  
Iteration 2:   log pseudolikelihood = -1855.3321  
Iteration 3:   log pseudolikelihood = -1855.1326  
Iteration 4:   log pseudolikelihood = -1855.1325  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     255.78
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1855.1325               Pseudo R2         =     0.1328

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_2 |   .4717276   .1178407     4.00   0.000     .2407639    .7026912
                RESTRICTsq_2 |  -.0439334    .015323    -2.87   0.004    -.0739659   -.0139009
                   PTS_Slag1 |   .6062909   .0977661     6.20   0.000     .4146729     .797909
                hrgroupslag1 |   .0009063   .0025781     0.35   0.725    -.0041467    .0059593
                  hrnewslag1 |   .1998701   .0553731     3.61   0.000     .0913408    .3083993
  protest_ClarkRegan_loglag1 |   .3073885   .0963181     3.19   0.001     .1186084    .4961685
      UCDP_armedConflictlag1 |   .3630916    .242395     1.50   0.134    -.1119939     .838177
         PR_freedomHouselag1 |   .7271115   .2279872     3.19   0.001     .2802647    1.173958
      PR_freedomHouselag1_sq |  -.0708555    .027364    -2.59   0.010     -.124488   -.0172231
   gdp_pc_constantUS2010lag1 |   .8300965   .3497794     2.37   0.018     .1445414    1.515652
gdp_pc_constantUS2010lag1_sq |  -.2842744   .1300612    -2.19   0.029    -.5391896   -.0293591
                   KOFGIlag1 |   .1042825   .0463386     2.25   0.024     .0134605    .1951046
                KOFGIlag1_sq |  -.0008953   .0004334    -2.07   0.039    -.0017448   -.0000458
              populationlag1 |   .1089059   .1746809     0.62   0.533    -.2334623    .4512741
                       _cons |  -6.308019   1.355819    -4.65   0.000    -8.965375   -3.650663
-----------------------------+----------------------------------------------------------------
                    /lnalpha |   .5565227   .1227496                       .315938    .7971074
-----------------------------+----------------------------------------------------------------
                       alpha |   1.744595   .2141483                      1.371545    2.219113
----------------------------------------------------------------------------------------------

.         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1855.133      16    3742.265   3824.359
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4l.doc", append ///
>         keep(RESTRICT_2 RESTRICTsq_2) ///
>         ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4l.doc
dir : seeout

. 
.         * Zero inflated negative binomial 
.         #delimit ;
delimiter now ;
.         zinb urgentActions RESTRICT_2 RESTRICTsq_2
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, 
>         inflate(fhbest fhworst UCDP_armedConflictlag1 deathpenalty urgentActionslag1) 
>         vce(cluster cowcode);

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1920.7697  
Iteration 2:   log pseudolikelihood = -1870.2721  (not concave)
Iteration 3:   log pseudolikelihood = -1796.4144  
Iteration 4:   log pseudolikelihood = -1781.2857  
Iteration 5:   log pseudolikelihood = -1780.4708  
Iteration 6:   log pseudolikelihood = -1780.4666  
Iteration 7:   log pseudolikelihood = -1780.4666  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     210.82
Log pseudolikelihood = -1780.467                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
                  RESTRICT_2 |   .3571338   .0964229     3.70   0.000     .1681484    .5461192
                RESTRICTsq_2 |  -.0356288   .0133513    -2.67   0.008    -.0617969   -.0094607
                   PTS_Slag1 |   .4575071   .0905832     5.05   0.000     .2799673    .6350469
                hrgroupslag1 |  -.0005596   .0022261    -0.25   0.802    -.0049227    .0038035
                  hrnewslag1 |   .1314492    .034349     3.83   0.000     .0641264    .1987719
  protest_ClarkRegan_loglag1 |   .2994305   .0781698     3.83   0.000     .1462206    .4526405
      UCDP_armedConflictlag1 |   .2825668   .2267468     1.25   0.213    -.1618487    .7269823
         PR_freedomHouselag1 |   .5167409   .2737821     1.89   0.059    -.0198623    1.053344
      PR_freedomHouselag1_sq |   -.055519   .0312301    -1.78   0.075     -.116729    .0056909
   gdp_pc_constantUS2010lag1 |   .7805937   .3597712     2.17   0.030     .0754551    1.485732
gdp_pc_constantUS2010lag1_sq |  -.2933216   .1382376    -2.12   0.034    -.5642623   -.0223809
                   KOFGIlag1 |   .0902866   .0424184     2.13   0.033     .0071481    .1734251
                KOFGIlag1_sq |  -.0008029   .0004013    -2.00   0.045    -.0015894   -.0000164
              populationlag1 |   .0930774   .1791998     0.52   0.603    -.2581477    .4443026
                       _cons |  -4.140172   1.273849    -3.25   0.001    -6.636871   -1.643473
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |    .554688   .5197093     1.07   0.286    -.4639235      1.5733
                     fhworst |  -1.812812   1.035054    -1.75   0.080     -3.84148    .2158562
      UCDP_armedConflictlag1 |  -.7652895   .4050087    -1.89   0.059    -1.559092    .0285129
                deathpenalty |   .1523472   .3080518     0.49   0.621    -.4514232    .7561175
           urgentActionslag1 |  -1.241272   .2269033    -5.47   0.000    -1.685994   -.7965495
                       _cons |   .3562605   .2560958     1.39   0.164    -.1456781     .858199
-----------------------------+----------------------------------------------------------------
                    /lnalpha |    -.10258   .1409467    -0.73   0.467    -.3788306    .1736705
-----------------------------+----------------------------------------------------------------
                       alpha |   .9025059   .1272053                      .6846616    1.189663
----------------------------------------------------------------------------------------------

.  // vuong ;
>         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1780.467      22    3604.933   3717.813
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4l.doc", append ///
>         keep(RESTRICT_2 RESTRICTsq_2) ///
>         ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4l.doc
dir : seeout

.         
.         
.         * GMM
.         xtset, clear

.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1
>         ( RESTRICT_2 RESTRICTsq_2 =
>                 RESTRICT_2lag2 RESTRICT_2lag3 RESTRICTsq_2lag2)
>                 , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.......x..........................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_2 |    .638581   .2507914     2.55   0.011     .1470389    1.130123
                RESTRICTsq_2 |  -.0626817   .0292942    -2.14   0.032    -.1200973   -.0052662
                   PTS_Slag1 |   .4705152   .1240822     3.79   0.000     .2273185    .7137118
                hrgroupslag1 |  -.0004452   .0028275    -0.16   0.875     -.005987    .0050966
                  hrnewslag1 |   .1277949   .0263168     4.86   0.000     .0762149    .1793749
  protest_ClarkRegan_loglag1 |   .2613372   .1037247     2.52   0.012     .0580405    .4646338
      UCDP_armedConflictlag1 |   .3403805   .2637221     1.29   0.197    -.1765053    .8572662
         PR_freedomHouselag1 |   .6830244   .3421772     2.00   0.046     .0123693    1.353679
      PR_freedomHouselag1_sq |  -.0794258   .0384238    -2.07   0.039    -.1547351   -.0041166
   gdp_pc_constantUS2010lag1 |   1.021053   .4700901     2.17   0.030     .0996931    1.942413
gdp_pc_constantUS2010lag1_sq |  -.4333324   .2253284    -1.92   0.054     -.874968    .0083032
                   KOFGIlag1 |   .0972683   .0782834     1.24   0.214    -.0561644     .250701
                KOFGIlag1_sq |  -.0009604   .0007281    -1.32   0.187    -.0023873    .0004666
              populationlag1 |  -.0171377   .2508084    -0.07   0.946    -.5087132    .4744378
                       _cons |  -4.779037   2.031046    -2.35   0.019    -8.759813   -.7982601
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_2 RESTRICTsq_2
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_2lag2 RESTRICT_2lag3 RESTRICTsq_2lag2
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

.         #delimit cr
delimiter now cr
.         
.         estat overid

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .420823 (p = 0.5165)

.         mat es_ic = r(J) 

.         matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .42082275

.         local J: display %4.1f es_ic[1,1]

.         outreg2 using ".\Tables\Appendix_TableS4l.doc", append ///
>         keep( RESTRICT_2 RESTRICTsq_2) ///
>         ctitle("Model 4") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Hansen's J, `J')
.\Tables\Appendix_TableS4l.doc
dir : seeout

. 
.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>                 gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>                 KOFGIlag1 KOFGIlag1_sq 
>                 hrgroupslag1 hrnewslag1 
>                 populationlag1  UCDP_armedConflictlag1 
>                 (RESTRICT_2 RESTRICTsq_2 
>                 PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
>                 protest_ClarkRegan_loglag1 = 
>                 RESTRICT_2lag2 RESTRICT_2lag3 RESTRICTsq_2lag2
>                 PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
>                 protest_ClarkRegan_loglag2 )
>         , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
.......x.....x....................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,248
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_2 |    .659341   .2735727     2.41   0.016     .1231484    1.195534
                RESTRICTsq_2 |  -.0650171   .0314279    -2.07   0.039    -.1266147   -.0034195
                   PTS_Slag1 |   .6336671   .1915924     3.31   0.001      .258153    1.009181
         PR_freedomHouselag1 |   .5535377   .4074798     1.36   0.174     -.245108    1.352183
      PR_freedomHouselag1_sq |  -.0700712   .0449387    -1.56   0.119    -.1581493     .018007
  protest_ClarkRegan_loglag1 |   .3511649   .1876948     1.87   0.061    -.0167102      .71904
   gdp_pc_constantUS2010lag1 |   1.043191   .4867284     2.14   0.032     .0892206    1.997161
gdp_pc_constantUS2010lag1_sq |  -.4058536   .2373125    -1.71   0.087    -.8709775    .0592703
                   KOFGIlag1 |   .0922593   .0779959     1.18   0.237    -.0606099    .2451284
                KOFGIlag1_sq |  -.0008739   .0007353    -1.19   0.235     -.002315    .0005672
                hrgroupslag1 |  -.0031953   .0037462    -0.85   0.394    -.0105376    .0041471
                  hrnewslag1 |    .110044   .0301238     3.65   0.000     .0510026    .1690855
              populationlag1 |   -.003768   .2564714    -0.01   0.988    -.5064426    .4989067
      UCDP_armedConflictlag1 |   .2084837    .314847     0.66   0.508    -.4086051    .8255725
                       _cons |  -4.861043   2.180687    -2.23   0.026    -9.135112   -.5869753
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_2 RESTRICTsq_2 PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1 RESTRICT_2lag2
               RESTRICT_2lag3 RESTRICTsq_2lag2 PTS_Slag2 PR_freedomHouselag2
               PR_freedomHouselag2_sq protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

.         #delimit cr
delimiter now cr
. 
.         estat overid 

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .558532 (p = 0.4549)

.         mat es_ic = r(J) 

.          matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .55853161

.         local J: display %4.1f es_ic[1,1]

.         outreg2 using ".\Tables\Appendix_TableS4l.doc", append ///
>          ctitle("Model 5") label  eqdrop(lnalpha) dec(3) ///
>          keep(RESTRICT_2 RESTRICTsq_2) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Hansen's J, `J')
.\Tables\Appendix_TableS4l.doc
dir : seeout

.         
.         
. 
.         
. ** Leave out: HARASS_AMOUNT_bilag1 - ARREST_bilag1 - KILLING_bilag1
. 
.         xtset cowcode YEAR
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

.         
.         * Negative binomial 
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_3 RESTRICTsq_3, vce(cluster cowcode) ;

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -5765.4996  
Iteration 1:   log pseudolikelihood = -5765.1483  
Iteration 2:   log pseudolikelihood = -5765.1483  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -3235.0249  
Iteration 1:   log pseudolikelihood = -2752.8167  
Iteration 2:   log pseudolikelihood = -2746.3213  
Iteration 3:   log pseudolikelihood = -2746.3144  
Iteration 4:   log pseudolikelihood = -2746.3144  

Fitting full model:

Iteration 0:   log pseudolikelihood = -2690.1039  
Iteration 1:   log pseudolikelihood = -2678.5161  
Iteration 2:   log pseudolikelihood = -2677.4224  
Iteration 3:   log pseudolikelihood = -2677.4196  
Iteration 4:   log pseudolikelihood = -2677.4196  

Negative binomial regression                    Number of obs     =      1,688
                                                Wald chi2(2)      =      38.40
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -2677.4196               Pseudo R2         =     0.0251

                               (Std. Err. adjusted for 171 clusters in cowcode)
-------------------------------------------------------------------------------
              |               Robust
urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
   RESTRICT_3 |   .8078135   .1359851     5.94   0.000     .5412876    1.074339
 RESTRICTsq_3 |  -.0740495   .0142533    -5.20   0.000    -.1019855   -.0461136
        _cons |    .069161   .1993907     0.35   0.729    -.3216377    .4599596
--------------+----------------------------------------------------------------
     /lnalpha |   1.381572   .1248275                      1.136915     1.62623
--------------+----------------------------------------------------------------
        alpha |   3.981156   .4969579                      3.117137    5.084668
-------------------------------------------------------------------------------

.         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,688 -2746.314   -2677.42       4    5362.839   5384.564
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         mat list es_ic

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1688  -2746.3144  -2677.4196           4   5362.8391   5384.5643

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4n.doc", replace ///
>         keep(RESTRICT_3 RESTRICTsq_3) ///
>         ctitle("Model 1") label eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4n.doc
dir : seeout

. 
.         * Negative binomial
.         #delimit ;
delimiter now ;
.         nbreg urgentActions RESTRICT_3 RESTRICTsq_3
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, vce(cluster cowcode);

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -12987.955  
Iteration 1:   log pseudolikelihood = -6675.8553  
Iteration 2:   log pseudolikelihood = -4348.6854  
Iteration 3:   log pseudolikelihood = -3096.5591  
Iteration 4:   log pseudolikelihood = -2996.0058  
Iteration 5:   log pseudolikelihood = -2993.1546  
Iteration 6:   log pseudolikelihood = -2993.1499  
Iteration 7:   log pseudolikelihood = -2993.1499  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2485.4691  
Iteration 1:   log pseudolikelihood = -2141.4633  
Iteration 2:   log pseudolikelihood = -2139.1792  
Iteration 3:   log pseudolikelihood = -2139.1783  
Iteration 4:   log pseudolikelihood = -2139.1783  

Fitting full model:

Iteration 0:   log pseudolikelihood =  -2014.275  
Iteration 1:   log pseudolikelihood = -2007.5158  
Iteration 2:   log pseudolikelihood = -1874.8498  
Iteration 3:   log pseudolikelihood = -1859.9779  
Iteration 4:   log pseudolikelihood =  -1859.849  
Iteration 5:   log pseudolikelihood = -1859.8489  

Negative binomial regression                    Number of obs     =      1,250
                                                Wald chi2(14)     =     277.75
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1859.8489               Pseudo R2         =     0.1306

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_3 |   .4134871   .1073097     3.85   0.000     .2031639    .6238102
                RESTRICTsq_3 |  -.0309745   .0115577    -2.68   0.007    -.0536272   -.0083219
                   PTS_Slag1 |   .6528843   .1000391     6.53   0.000     .4568111    .8489574
                hrgroupslag1 |   .0033971   .0028875     1.18   0.239    -.0022623    .0090565
                  hrnewslag1 |   .1995864    .057513     3.47   0.001      .086863    .3123098
  protest_ClarkRegan_loglag1 |   .3363767   .0907972     3.70   0.000     .1584175    .5143359
      UCDP_armedConflictlag1 |   .4600516    .290979     1.58   0.114    -.1102568     1.03036
         PR_freedomHouselag1 |   .7346196   .2506882     2.93   0.003     .2432798    1.225959
      PR_freedomHouselag1_sq |  -.0773836   .0298584    -2.59   0.010    -.1359051   -.0188621
   gdp_pc_constantUS2010lag1 |   .8137899   .3687269     2.21   0.027     .0910984    1.536481
gdp_pc_constantUS2010lag1_sq |  -.2766895   .1366903    -2.02   0.043    -.5445975   -.0087814
                   KOFGIlag1 |   .1129097   .0479388     2.36   0.019     .0189513    .2068681
                KOFGIlag1_sq |   -.001016   .0004578    -2.22   0.026    -.0019132   -.0001188
              populationlag1 |    .147726   .1735699     0.85   0.395    -.1924647    .4879166
                       _cons |  -6.615608   1.313217    -5.04   0.000    -9.189465   -4.041751
-----------------------------+----------------------------------------------------------------
                    /lnalpha |    .578605   .1357264                       .312586    .8446239
-----------------------------+----------------------------------------------------------------
                       alpha |   1.783549   .2420747                      1.366956    2.327102
----------------------------------------------------------------------------------------------

.         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250 -2139.178  -1859.849      16    3751.698   3833.792
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4n.doc", append ///
>         keep(RESTRICT_3 RESTRICTsq_3) ///
>         ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4n.doc
dir : seeout

. 
.         * Zero inflated negative binomial 
.         #delimit ;
delimiter now ;
.         zinb urgentActions RESTRICT_3 RESTRICTsq_3
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1, 
>         inflate(fhbest fhworst UCDP_armedConflictlag1 deathpenalty urgentActionslag1) 
>         vce(cluster cowcode);

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2328.2805  
Iteration 1:   log pseudolikelihood = -2168.6869  
Iteration 2:   log pseudolikelihood = -2016.6097  
Iteration 3:   log pseudolikelihood = -1971.0267  
Iteration 4:   log pseudolikelihood = -1948.4438  
Iteration 5:   log pseudolikelihood = -1944.1592  
Iteration 6:   log pseudolikelihood = -1943.8511  
Iteration 7:   log pseudolikelihood = -1943.8495  
Iteration 8:   log pseudolikelihood = -1943.8495  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1943.8495  
Iteration 1:   log pseudolikelihood = -1883.7264  
Iteration 2:   log pseudolikelihood = -1830.7427  (not concave)
Iteration 3:   log pseudolikelihood = -1795.9205  
Iteration 4:   log pseudolikelihood = -1785.6539  
Iteration 5:   log pseudolikelihood = -1785.3007  
Iteration 6:   log pseudolikelihood = -1785.2994  
Iteration 7:   log pseudolikelihood = -1785.2994  

Zero-inflated negative binomial regression      Number of obs     =      1,250
                                                Nonzero obs       =        537
                                                Zero obs          =        713

Inflation model      = logit                    Wald chi2(14)     =     238.92
Log pseudolikelihood = -1785.299                Prob > chi2       =     0.0000

                                              (Std. Err. adjusted for 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |               Robust
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
urgentActions                |
                  RESTRICT_3 |   .2957745   .1031145     2.87   0.004     .0936738    .4978752
                RESTRICTsq_3 |  -.0236839    .011405    -2.08   0.038    -.0460373   -.0013304
                   PTS_Slag1 |   .4911791   .0928055     5.29   0.000     .3092836    .6730745
                hrgroupslag1 |   .0015202    .002479     0.61   0.540    -.0033385     .006379
                  hrnewslag1 |   .1332826   .0366146     3.64   0.000     .0615193     .205046
  protest_ClarkRegan_loglag1 |   .3199537   .0737792     4.34   0.000     .1753491    .4645583
      UCDP_armedConflictlag1 |   .3380337   .2632511     1.28   0.199    -.1779291    .8539965
         PR_freedomHouselag1 |   .5247734   .3024591     1.74   0.083    -.0680356    1.117582
      PR_freedomHouselag1_sq |  -.0606481   .0344951    -1.76   0.079    -.1282573    .0069612
   gdp_pc_constantUS2010lag1 |   .7529357   .3605763     2.09   0.037     .0462191    1.459652
gdp_pc_constantUS2010lag1_sq |   -.277299   .1374632    -2.02   0.044    -.5467218   -.0078761
                   KOFGIlag1 |   .0943735   .0434033     2.17   0.030     .0093045    .1794424
                KOFGIlag1_sq |  -.0008794   .0004129    -2.13   0.033    -.0016887   -.0000701
              populationlag1 |   .1247576   .1781321     0.70   0.484    -.2243749      .47389
                       _cons |  -4.292146   1.238784    -3.46   0.001    -6.720117   -1.864174
-----------------------------+----------------------------------------------------------------
inflate                      |
                      fhbest |   .5577396   .5285352     1.06   0.291    -.4781704     1.59365
                     fhworst |  -1.972655   1.170024    -1.69   0.092    -4.265859    .3205498
      UCDP_armedConflictlag1 |  -.7619004   .4257642    -1.79   0.074    -1.596383     .072582
                deathpenalty |   .1689289   .3100729     0.54   0.586    -.4388029    .7766607
           urgentActionslag1 |  -1.255836   .2253947    -5.57   0.000    -1.697601   -.8140703
                       _cons |   .3514092   .2571423     1.37   0.172    -.1525804    .8553987
-----------------------------+----------------------------------------------------------------
                    /lnalpha |  -.0725832   .1500802    -0.48   0.629     -.366735    .2215686
-----------------------------+----------------------------------------------------------------
                       alpha |   .9299884   .1395729                      .6929933    1.248033
----------------------------------------------------------------------------------------------

.  // vuong ;
>         #delimit cr
delimiter now cr
. 
.         estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,250  -1943.85  -1785.299      22    3614.599   3727.479
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

.         mat es_ic = r(S)

.         local AIC: display %4.1f es_ic[1,5]

.         local BIC: display %4.1f es_ic[1,6]

.         local LL: display %4.1f es_ic[1,3]

.         outreg2 using ".\Tables\Appendix_TableS4n.doc", append ///
>         keep(RESTRICT_3 RESTRICTsq_3) ///
>         ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS4n.doc
dir : seeout

.                 
.         * GMM
.         xtset, clear

.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>         PTS_Slag1 hrgroupslag1 hrnewslag1 
>         protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
>         PR_freedomHouselag1 PR_freedomHouselag1_sq 
>         gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>         KOFGIlag1 KOFGIlag1_sq populationlag1
>         ( RESTRICT_3 RESTRICTsq_3 =
>                 RESTRICT_3lag2 RESTRICT_3lag3 RESTRICTsq_3lag2)
>                 , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,245
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_3 |   .4264162    .142159     3.00   0.003     .1477897    .7050427
                RESTRICTsq_3 |  -.0345277   .0184547    -1.87   0.061    -.0706982    .0016429
                   PTS_Slag1 |   .5520275   .1193773     4.62   0.000     .3180523    .7860026
                hrgroupslag1 |   .0014964   .0027856     0.54   0.591    -.0039633    .0069562
                  hrnewslag1 |   .1373331   .0280354     4.90   0.000     .0823848    .1922815
  protest_ClarkRegan_loglag1 |   .3044979   .1007329     3.02   0.003     .1070651    .5019307
      UCDP_armedConflictlag1 |   .4434263   .2574144     1.72   0.085    -.0610967    .9479492
         PR_freedomHouselag1 |   .6880276   .3505489     1.96   0.050     .0009644    1.375091
      PR_freedomHouselag1_sq |  -.0807842   .0423033    -1.91   0.056    -.1636972    .0021289
   gdp_pc_constantUS2010lag1 |   1.043181   .5171613     2.02   0.044     .0295636    2.056799
gdp_pc_constantUS2010lag1_sq |  -.4695625   .2655147    -1.77   0.077    -.9899617    .0508366
                   KOFGIlag1 |   .1049549   .0833261     1.26   0.208    -.0583613    .2682711
                KOFGIlag1_sq |  -.0010652    .000776    -1.37   0.170    -.0025861    .0004558
              populationlag1 |   .0036404   .2474319     0.01   0.988    -.4813173     .488598
                       _cons |  -5.066372   2.087618    -2.43   0.015    -9.158027   -.9747164
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_3 RESTRICTsq_3
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 RESTRICT_3lag2 RESTRICT_3lag3 RESTRICTsq_3lag2

.         #delimit cr
delimiter now cr
.         
.         estat overid

  Test of overidentifying restriction:

  Hansen's J chi2(1) =  .05228 (p = 0.8191)

.         mat es_ic = r(J) 

.         matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .05227955

.         local J: display %4.1f es_ic[1,1]

.         outreg2 using ".\Tables\Appendix_TableS4n.doc", append ///
>         keep( RESTRICT_3 RESTRICTsq_3) ///
>         ctitle("Model 4") label  eqdrop(lnalpha) dec(3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Hansen's J, `J')
.\Tables\Appendix_TableS4n.doc
dir : seeout

. 
. 
.         #delimit ;
delimiter now ;
.         ivpoisson gmm urgentActions
>                 gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
>                 KOFGIlag1 KOFGIlag1_sq 
>                 hrgroupslag1 hrnewslag1 
>                 populationlag1  UCDP_armedConflictlag1 
>                 (RESTRICT_3 RESTRICTsq_3 
>                 PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
>                 protest_ClarkRegan_loglag1 = 
>                 RESTRICT_3lag2 RESTRICT_3lag3 RESTRICTsq_3lag2
>                 PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
>                 protest_ClarkRegan_loglag2 )
>         , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50

Exponential mean model with endogenous regressors

Number of parameters =  15                         Number of obs  =      1,245
Number of moments    =  16
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 147 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
                  RESTRICT_3 |   .4685827   .1461897     3.21   0.001     .1820561    .7551092
                RESTRICTsq_3 |  -.0380995   .0182753    -2.08   0.037    -.0739184   -.0022805
                   PTS_Slag1 |   .7408423   .1987382     3.73   0.000     .3513227    1.130362
         PR_freedomHouselag1 |   .6189317   .4214021     1.47   0.142    -.2070012    1.444865
      PR_freedomHouselag1_sq |  -.0789508   .0517843    -1.52   0.127    -.1804462    .0225446
  protest_ClarkRegan_loglag1 |   .3893033   .1645745     2.37   0.018     .0667432    .7118634
   gdp_pc_constantUS2010lag1 |   1.063644   .5232863     2.03   0.042     .0380217    2.089267
gdp_pc_constantUS2010lag1_sq |  -.4408595   .2707433    -1.63   0.103    -.9715065    .0897876
                   KOFGIlag1 |   .0954658   .0843613     1.13   0.258    -.0698793    .2608109
                KOFGIlag1_sq |  -.0009277   .0007902    -1.17   0.240    -.0024765    .0006212
                hrgroupslag1 |  -.0011695   .0035639    -0.33   0.743    -.0081546    .0058156
                  hrnewslag1 |   .1201633   .0326342     3.68   0.000     .0562015     .184125
              populationlag1 |   .0151149   .2547998     0.06   0.953    -.4842836    .5145134
      UCDP_armedConflictlag1 |   .2711726   .3046782     0.89   0.373    -.3259857    .8683308
                       _cons |  -5.258391   2.195667    -2.39   0.017     -9.56182    -.954962
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_3 RESTRICTsq_3 PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1 RESTRICT_3lag2
               RESTRICT_3lag3 RESTRICTsq_3lag2 PTS_Slag2 PR_freedomHouselag2
               PR_freedomHouselag2_sq protest_ClarkRegan_loglag2

.         #delimit cr
delimiter now cr
. 
.         estat overid 

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .091504 (p = 0.7623)

.         mat es_ic = r(J) 

.          matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .09150411

.         local J: display %4.1f es_ic[1,1]

.         outreg2 using ".\Tables\Appendix_TableS4n.doc", append ///
>          ctitle("Model 5") label  eqdrop(lnalpha) dec(3) ///
>          keep(RESTRICT_3 RESTRICTsq_3) ///
>         alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
>         addtext(Hansen's J, `J')
.\Tables\Appendix_TableS4n.doc
dir : seeout

.         
.         
.         
.         
.         
. ****************************************************************************
. **** Appendix S5: Regime Types and Restrictions
. ****************************************************************************
.         
. gen polity2lag1_sq = polity2lag1*polity2lag1
(2,312 missing values generated)

. 
. * Model 2: Negative binomial with robust se
. #delimit ;
delimiter now ;
. nbreg urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1 
> polity2lag1 polity2lag1_sq, vce(cluster cowcode);

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -11995.056  
Iteration 1:   log pseudolikelihood = -6113.0168  
Iteration 2:   log pseudolikelihood = -4034.4166  
Iteration 3:   log pseudolikelihood = -2851.8307  
Iteration 4:   log pseudolikelihood = -2741.3607  
Iteration 5:   log pseudolikelihood = -2739.2778  
Iteration 6:   log pseudolikelihood = -2739.2765  
Iteration 7:   log pseudolikelihood = -2739.2765  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2442.2253  
Iteration 1:   log pseudolikelihood = -2095.5218  
Iteration 2:   log pseudolikelihood = -2093.3869  
Iteration 3:   log pseudolikelihood =  -2093.386  
Iteration 4:   log pseudolikelihood =  -2093.386  

Fitting full model:

Iteration 0:   log pseudolikelihood =  -1963.441  
Iteration 1:   log pseudolikelihood = -1796.2654  
Iteration 2:   log pseudolikelihood = -1779.4396  
Iteration 3:   log pseudolikelihood = -1779.1946  
Iteration 4:   log pseudolikelihood = -1779.1945  

Negative binomial regression                    Number of obs     =      1,227
                                                Wald chi2(16)     =     325.70
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -1779.1945               Pseudo R2         =     0.1501

                                                                 (Std. Err. adjusted for 145 clusters in cowcode)
-----------------------------------------------------------------------------------------------------------------
                                                |               Robust
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |   .5485389   .1104054     4.97   0.000     .3321482    .7649295
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0448213   .0111427    -4.02   0.000    -.0666605    -.022982
                                                |
                                      PTS_Slag1 |   .6124748   .0934625     6.55   0.000     .4292917     .795658
                                   hrgroupslag1 |   .0021059   .0027558     0.76   0.445    -.0032954    .0075072
                                     hrnewslag1 |   .1887538   .0588768     3.21   0.001     .0733574    .3041502
                     protest_ClarkRegan_loglag1 |   .2974406   .0943706     3.15   0.002     .1124777    .4824035
                         UCDP_armedConflictlag1 |   .3730416   .2240204     1.67   0.096    -.0660304    .8121136
                            PR_freedomHouselag1 |   1.226253   .2946929     4.16   0.000     .6486655    1.803841
                         PR_freedomHouselag1_sq |  -.1198104   .0348002    -3.44   0.001    -.1880176   -.0516032
                      gdp_pc_constantUS2010lag1 |   .5278902   .3320285     1.59   0.112    -.1228736    1.178654
                   gdp_pc_constantUS2010lag1_sq |  -.2235532   .1162702    -1.92   0.055    -.4514387    .0043323
                                      KOFGIlag1 |   .0589374   .0466868     1.26   0.207    -.0325671     .150442
                                   KOFGIlag1_sq |  -.0004545   .0004612    -0.99   0.324    -.0013584    .0004494
                                 populationlag1 |   .1531288   .1710599     0.90   0.371    -.1821424       .4884
                                    polity2lag1 |   .0130957   .0339218     0.39   0.699    -.0533898    .0795811
                                 polity2lag1_sq |     .01517   .0040041     3.79   0.000     .0073222    .0230178
                                          _cons |  -7.447957   1.279829    -5.82   0.000    -9.956375   -4.939538
------------------------------------------------+----------------------------------------------------------------
                                       /lnalpha |   .4496934   .1241188                       .206425    .6929619
------------------------------------------------+----------------------------------------------------------------
                                          alpha |   1.567831   .1945974                      1.229276    1.999629
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,227 -2093.386  -1779.195      18    3594.389   3686.411
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. mat list es_ic

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        1227   -2093.386  -1779.1945          18   3594.3891   3686.4109

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS5.doc", replace ///
>  ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS5.doc
dir : seeout

. 
. * Model 3: Zero inflated negative binomial 
. set seed 2

. #delimit ;
delimiter now ;
. zinb urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1
> polity2lag1 polity2lag1_sq, 
> inflate(fhbest fhworst UCDP_armedConflictlag1 deathpenalty urgentActionslag1) 
> vce(cluster cowcode);

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -2281.1071  
Iteration 1:   log pseudolikelihood = -2134.2744  
Iteration 2:   log pseudolikelihood = -1973.2032  
Iteration 3:   log pseudolikelihood =  -1927.256  
Iteration 4:   log pseudolikelihood = -1904.7842  
Iteration 5:   log pseudolikelihood = -1900.4782  
Iteration 6:   log pseudolikelihood = -1900.1885  
Iteration 7:   log pseudolikelihood = -1900.1871  
Iteration 8:   log pseudolikelihood = -1900.1871  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1900.1871  
Iteration 1:   log pseudolikelihood = -1800.0967  
Iteration 2:   log pseudolikelihood = -1742.0054  (not concave)
Iteration 3:   log pseudolikelihood = -1717.0038  
Iteration 4:   log pseudolikelihood = -1713.5409  
Iteration 5:   log pseudolikelihood = -1713.4913  
Iteration 6:   log pseudolikelihood = -1713.4913  

Zero-inflated negative binomial regression      Number of obs     =      1,227
                                                Nonzero obs       =        523
                                                Zero obs          =        704

Inflation model      = logit                    Wald chi2(16)     =     292.74
Log pseudolikelihood = -1713.491                Prob > chi2       =     0.0000

                                                                 (Std. Err. adjusted for 145 clusters in cowcode)
-----------------------------------------------------------------------------------------------------------------
                                                |               Robust
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
urgentActions                                   |
                          RESTRICT_COUNTdomlag1 |   .4418478   .0952863     4.64   0.000     .2550901    .6286055
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |   -.038296   .0097678    -3.92   0.000    -.0574406   -.0191515
                                                |
                                      PTS_Slag1 |   .5113748   .0870218     5.88   0.000     .3408152    .6819345
                                   hrgroupslag1 |   .0007197   .0025573     0.28   0.778    -.0042926     .005732
                                     hrnewslag1 |   .1276542   .0376273     3.39   0.001      .053906    .2014023
                     protest_ClarkRegan_loglag1 |   .2867779   .0748882     3.83   0.000     .1399998    .4335561
                         UCDP_armedConflictlag1 |   .2086349   .2068127     1.01   0.313    -.1967105    .6139803
                            PR_freedomHouselag1 |   .8613424   .3503955     2.46   0.014     .1745799    1.548105
                         PR_freedomHouselag1_sq |  -.0946286   .0397671    -2.38   0.017    -.1725706   -.0166866
                      gdp_pc_constantUS2010lag1 |   .4523232   .3310732     1.37   0.172    -.1965682    1.101215
                   gdp_pc_constantUS2010lag1_sq |   -.219081   .1198361    -1.83   0.068    -.4539555    .0157935
                                      KOFGIlag1 |   .0434313   .0437822     0.99   0.321    -.0423803    .1292428
                                   KOFGIlag1_sq |   -.000338   .0004356    -0.78   0.438    -.0011917    .0005157
                                 populationlag1 |    .078894   .1743787     0.45   0.651    -.2628819      .42067
                                    polity2lag1 |   -.012092   .0355336    -0.34   0.734    -.0817365    .0575525
                                 polity2lag1_sq |   .0112132   .0040804     2.75   0.006     .0032157    .0192106
                                          _cons |  -4.724053   1.331611    -3.55   0.000    -7.333963   -2.114144
------------------------------------------------+----------------------------------------------------------------
inflate                                         |
                                         fhbest |   .5096672   .5347023     0.95   0.340    -.5383301    1.557665
                                        fhworst |  -1.839139   1.168579    -1.57   0.116    -4.129512    .4512335
                         UCDP_armedConflictlag1 |  -.7108921   .4068089    -1.75   0.081    -1.508223    .0864387
                                   deathpenalty |   .0546069   .3186475     0.17   0.864    -.5699307    .6791444
                              urgentActionslag1 |  -1.198349   .2183026    -5.49   0.000    -1.626214   -.7704833
                                          _cons |   .3423955   .2612506     1.31   0.190    -.1696463    .8544374
------------------------------------------------+----------------------------------------------------------------
                                       /lnalpha |  -.1841144    .137181    -1.34   0.180    -.4529842    .0847554
------------------------------------------------+----------------------------------------------------------------
                                          alpha |   .8318407   .1141127                      .6357282    1.088451
-----------------------------------------------------------------------------------------------------------------

.  // vuong ;
> #delimit cr
delimiter now cr
. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,227 -1900.187  -1713.491      24    3474.983   3597.678
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS5.doc", append ///
>  ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS5.doc
dir : seeout

. 
. 
. ** Model 4: GMM (with 2 EEVs)
. xtset, clear

. #delimit ;
delimiter now ;
. ivpoisson gmm urgentActions
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1
> polity2lag1 polity2lag1_sq
> ( RESTRICT_COUNTdomlag1 RESTRICT_COUNTdomlag1_sq = 
>         RESTRICT_COUNTdomlag2 RESTRICT_COUNTdomlag3
>         RESTRICT_COUNTdomlag2_sq)
>         , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x....................x......    50

Exponential mean model with endogenous regressors

Number of parameters =  17                         Number of obs  =      1,225
Number of moments    =  18
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 145 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
       RESTRICT_COUNTdomlag1 |    .712107   .2425311     2.94   0.003     .2367547    1.187459
    RESTRICT_COUNTdomlag1_sq |  -.0633901    .027082    -2.34   0.019    -.1164699   -.0103103
                   PTS_Slag1 |   .5950642   .1394801     4.27   0.000     .3216883    .8684401
                hrgroupslag1 |  -.0021344    .002968    -0.72   0.472    -.0079515    .0036827
                  hrnewslag1 |   .1113622   .0427384     2.61   0.009     .0275965    .1951279
  protest_ClarkRegan_loglag1 |   .2870492    .090906     3.16   0.002     .1088766    .4652217
      UCDP_armedConflictlag1 |   .2478289   .2168138     1.14   0.253    -.1771184    .6727761
         PR_freedomHouselag1 |   .8703364   .3399057     2.56   0.010     .2041335    1.536539
      PR_freedomHouselag1_sq |  -.1063235   .0446043    -2.38   0.017    -.1937464   -.0189007
   gdp_pc_constantUS2010lag1 |   .7553842   .5502721     1.37   0.170    -.3231293    1.833898
gdp_pc_constantUS2010lag1_sq |  -.4397759   .3415283    -1.29   0.198    -1.109159    .2296073
                   KOFGIlag1 |   .0325291   .0793494     0.41   0.682    -.1229929    .1880512
                KOFGIlag1_sq |  -.0002572   .0007817    -0.33   0.742    -.0017893     .001275
              populationlag1 |   .1220272   .2367436     0.52   0.606    -.3419817    .5860361
                 polity2lag1 |  -.0393807   .0582826    -0.68   0.499    -.1536125     .074851
              polity2lag1_sq |   .0147387   .0062958     2.34   0.019     .0023992    .0270782
                       _cons |  -4.913342   2.296283    -2.14   0.032    -9.413973   -.4127112
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_COUNTdomlag1 RESTRICT_COUNTdomlag1_sq
Instruments:   PTS_Slag1 hrgroupslag1 hrnewslag1 protest_ClarkRegan_loglag1
               UCDP_armedConflictlag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
               gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               populationlag1 polity2lag1 polity2lag1_sq RESTRICT_COUNTdomlag2
               RESTRICT_COUNTdomlag3 RESTRICT_COUNTdomlag2_sq
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

. #delimit cr
delimiter now cr
. estat overid

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .050149 (p = 0.8228)

. 
. mat es_ic = r(J) 

. matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .05014865

. local J: display %4.1f es_ic[1,1]

. outreg2 using ".\Tables\Appendix_TableS5.doc", append ///
>  ctitle("Model 4") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Hansen's J, `J')
.\Tables\Appendix_TableS5.doc
dir : seeout

. 
. 
. ** Model 5: GMM (with all EEVs)
. xtset, clear

. #delimit ;
delimiter now ;
. ivpoisson gmm urgentActions
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq 
> hrgroupslag1 hrnewslag1 
> populationlag1  UCDP_armedConflictlag1 
> polity2lag1 polity2lag1_sq
> (RESTRICT_COUNTdomlag1 RESTRICT_COUNTdomlag1_sq 
> PTS_Slag1 PR_freedomHouselag1 PR_freedomHouselag1_sq
> protest_ClarkRegan_loglag1 = 
>         RESTRICT_COUNTdomlag2 RESTRICT_COUNTdomlag3
>         RESTRICT_COUNTdomlag2_sq
>         PTS_Slag2 
>         PR_freedomHouselag2
>         PR_freedomHouselag2_sq 
>         protest_ClarkRegan_loglag2 )
> , twostep vce(boot, reps(50) cl(cowcode) seed(1)) ;
(running ivpoisson on estimation sample)

Bootstrap replications (50)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
......................x.x.........................    50

Exponential mean model with endogenous regressors

Number of parameters =  17                         Number of obs  =      1,225
Number of moments    =  18
Initial weight matrix: Unadjusted
GMM weight matrix:     Robust

                                               (Replications based on 145 clusters in cowcode)
----------------------------------------------------------------------------------------------
                             |   Observed   Bootstrap                         Normal-based
               urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-----------------------------+----------------------------------------------------------------
       RESTRICT_COUNTdomlag1 |    .764325   .2691116     2.84   0.005     .2368759    1.291774
    RESTRICT_COUNTdomlag1_sq |  -.0674444   .0282465    -2.39   0.017    -.1228065   -.0120822
                   PTS_Slag1 |   .8521553   .2447646     3.48   0.000     .3724254    1.331885
         PR_freedomHouselag1 |   .5454695   .5010209     1.09   0.276    -.4365135    1.527453
      PR_freedomHouselag1_sq |  -.0869548   .0509361    -1.71   0.088    -.1867877    .0128782
  protest_ClarkRegan_loglag1 |   .4207002   .1985175     2.12   0.034      .031613    .8097875
   gdp_pc_constantUS2010lag1 |   .7969236    .526894     1.51   0.130    -.2357697    1.829617
gdp_pc_constantUS2010lag1_sq |  -.4394587   .3623641    -1.21   0.225    -1.149679    .2707618
                   KOFGIlag1 |   .0207156   .0851638     0.24   0.808    -.1462025    .1876337
                KOFGIlag1_sq |  -.0000587   .0008721    -0.07   0.946     -.001768    .0016506
                hrgroupslag1 |   -.005759   .0042329    -1.36   0.174    -.0140552    .0025373
                  hrnewslag1 |   .0890066   .0490044     1.82   0.069    -.0070402    .1850534
              populationlag1 |   .1463155   .2429636     0.60   0.547    -.3298844    .6225154
      UCDP_armedConflictlag1 |   .0817075   .2603057     0.31   0.754    -.4284823    .5918974
                 polity2lag1 |   -.069475    .080642    -0.86   0.389    -.2275305    .0885805
              polity2lag1_sq |   .0126367    .006337     1.99   0.046     .0002164    .0250569
                       _cons |  -4.562349   2.736899    -1.67   0.096    -9.926574    .8018752
----------------------------------------------------------------------------------------------
Instrumented:  RESTRICT_COUNTdomlag1 RESTRICT_COUNTdomlag1_sq PTS_Slag1 PR_freedomHouselag1
               PR_freedomHouselag1_sq protest_ClarkRegan_loglag1
Instruments:   gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq KOFGIlag1 KOFGIlag1_sq
               hrgroupslag1 hrnewslag1 populationlag1 UCDP_armedConflictlag1 polity2lag1
               polity2lag1_sq RESTRICT_COUNTdomlag2 RESTRICT_COUNTdomlag3
               RESTRICT_COUNTdomlag2_sq PTS_Slag2 PR_freedomHouselag2 PR_freedomHouselag2_sq
               protest_ClarkRegan_loglag2
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

. #delimit cr
delimiter now cr
. estat overid 

  Test of overidentifying restriction:

  Hansen's J chi2(1) = .030247 (p = 0.8619)

. 
. mat es_ic = r(J) 

.  matrix list es_ic

symmetric es_ic[1,1]
           c1
r1  .03024688

. local J: display %4.1f es_ic[1,1]

. outreg2 using ".\Tables\Appendix_TableS5.doc", append ///
>  ctitle("Model 5") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Hansen's J, `J')
.\Tables\Appendix_TableS5.doc
dir : seeout

. 
. 
. 
. ***************************************************************************************
. *** Appendix S6: Different operationalization of shaming using Murdie and David's (2012) data.
. ***************************************************************************************
. 
. * Model 1: Negative binomial with robust standard errors
. #delimit ;
delimiter now ;
. nbreg shamingINGO c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1, vce(cluster cowcode) ;

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -1817.2328  
Iteration 1:   log pseudolikelihood = -1817.2311  
Iteration 2:   log pseudolikelihood = -1817.2311  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -1439.9908  (not concave)
Iteration 1:   log pseudolikelihood = -1085.4146  
Iteration 2:   log pseudolikelihood = -1072.3255  
Iteration 3:   log pseudolikelihood = -1072.3228  
Iteration 4:   log pseudolikelihood = -1072.3228  

Fitting full model:

Iteration 0:   log pseudolikelihood = -1064.3926  
Iteration 1:   log pseudolikelihood = -1063.8185  
Iteration 2:   log pseudolikelihood = -1063.8142  
Iteration 3:   log pseudolikelihood = -1063.8142  

Negative binomial regression                    Number of obs     =      2,121
                                                Wald chi2(2)      =       8.17
Dispersion           = mean                     Prob > chi2       =     0.0168
Log pseudolikelihood = -1063.8142               Pseudo R2         =     0.0079

                                                                 (Std. Err. adjusted for 170 clusters in cowcode)
-----------------------------------------------------------------------------------------------------------------
                                                |               Robust
                                    shamingINGO |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |    .402319   .1437282     2.80   0.005     .1206169    .6840211
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0359292   .0144932    -2.48   0.013    -.0643353    -.007523
                                                |
                                          _cons |  -1.627603   .2487385    -6.54   0.000    -2.115122   -1.140085
------------------------------------------------+----------------------------------------------------------------
                                       /lnalpha |    2.75648   .1348672                      2.492145    3.020814
------------------------------------------------+----------------------------------------------------------------
                                          alpha |   15.74432   2.123392                      12.08717    20.50799
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      2,121 -1072.323  -1063.814       4    2135.628   2158.267
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. mat list es_ic

es_ic[1,6]
            N         ll0          ll          df         AIC         BIC
.        2121  -1072.3228  -1063.8142           4   2135.6285    2158.267

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS6.doc", replace ///
>  ctitle("Model 1") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS6.doc
dir : seeout

. 
. 
. margins, at(RESTRICT_COUNTdomlag1 = (0(1)10) )

Adjusted predictions                            Number of obs     =      2,121
Model VCE    : Robust

Expression   : Predicted number of events, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           1

3._at        : RESTRI~mlag1    =           2

4._at        : RESTRI~mlag1    =           3

5._at        : RESTRI~mlag1    =           4

6._at        : RESTRI~mlag1    =           5

7._at        : RESTRI~mlag1    =           6

8._at        : RESTRI~mlag1    =           7

9._at        : RESTRI~mlag1    =           8

10._at       : RESTRI~mlag1    =           9

11._at       : RESTRI~mlag1    =          10

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .1963997   .0488522     4.02   0.000     .1006512    .2921482
          2  |     .28331   .0476591     5.94   0.000     .1898999    .3767202
          3  |    .380343   .0596136     6.38   0.000     .2635024    .4971835
          4  |   .4752051   .0882461     5.38   0.000      .302246    .6481642
          5  |   .5525596   .1196737     4.62   0.000     .3180036    .7871157
          6  |   .5979564   .1411571     4.24   0.000     .3212936    .8746192
          7  |   .6022159   .1473645     4.09   0.000     .3133868     .891045
          8  |   .5644523   .1420606     3.97   0.000     .2860186     .842886
          9  |   .4923734   .1349602     3.65   0.000     .2278563    .7568905
         10  |   .3997185   .1321534     3.02   0.002     .1407026    .6587344
         11  |   .3019995    .129884     2.33   0.020     .0474316    .5565675
------------------------------------------------------------------------------

. marginsplot, recast(line) recastci(rarea)  ///
> plotopt(color(gs0) lwidth(1) ) ///
> ciopt(color(gs6) fintensity(10) lcolor(gs16) ) ///
> xtitle("Count of restriction types", size(large)) ///
> ytitle("Predicted number of media shaming events", size(large)) ///
> title("Model 1", size(large)) ///
> scheme(s1mono)

  Variables that uniquely identify margins: RESTRICT_COUNTdomlag1

. graph export ".\Figures\Appendix_FigureS6_1.png", replace
(note: file .\Figures\Appendix_FigureS6_1.png not found)
(file .\Figures\Appendix_FigureS6_1.png written in PNG format)

. 
. 
. * Negative binomial with robust se
. #delimit ;
delimiter now ;
. nbreg shamingINGO c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1, vce(cluster cowcode);

Fitting Poisson model:

Iteration 0:   log pseudolikelihood = -2607.3617  
Iteration 1:   log pseudolikelihood = -1638.9816  
Iteration 2:   log pseudolikelihood = -1337.7112  
Iteration 3:   log pseudolikelihood = -1000.6364  
Iteration 4:   log pseudolikelihood = -977.62335  
Iteration 5:   log pseudolikelihood =  -975.5445  
Iteration 6:   log pseudolikelihood = -975.53706  
Iteration 7:   log pseudolikelihood = -975.53706  

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -865.74419  (not concave)
Iteration 1:   log pseudolikelihood = -651.25741  
Iteration 2:   log pseudolikelihood = -640.74935  
Iteration 3:   log pseudolikelihood = -640.74668  
Iteration 4:   log pseudolikelihood = -640.74668  

Fitting full model:

Iteration 0:   log pseudolikelihood = -620.80607  
Iteration 1:   log pseudolikelihood = -607.03594  
Iteration 2:   log pseudolikelihood = -606.30314  
Iteration 3:   log pseudolikelihood = -606.30061  
Iteration 4:   log pseudolikelihood = -606.30061  

Negative binomial regression                    Number of obs     =      1,210
                                                Wald chi2(14)     =      61.94
Dispersion           = mean                     Prob > chi2       =     0.0000
Log pseudolikelihood = -606.30061               Pseudo R2         =     0.0538

                                                                 (Std. Err. adjusted for 147 clusters in cowcode)
-----------------------------------------------------------------------------------------------------------------
                                                |               Robust
                                    shamingINGO |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |    .267905    .160457     1.67   0.095    -.0465849    .5823949
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0360068   .0173004    -2.08   0.037    -.0699149   -.0020986
                                                |
                                      PTS_Slag1 |   .4394753   .1803703     2.44   0.015      .085956    .7929945
                                   hrgroupslag1 |  -.0042689   .0041385    -1.03   0.302    -.0123802    .0038424
                                     hrnewslag1 |   .3857148    .126141     3.06   0.002     .1384831    .6329465
                     protest_ClarkRegan_loglag1 |    .278437   .1569482     1.77   0.076    -.0291758    .5860498
                         UCDP_armedConflictlag1 |   .2394479   .4598577     0.52   0.603    -.6618565    1.140752
                            PR_freedomHouselag1 |   -.134522   .5087844    -0.26   0.791    -1.131721    .8626771
                         PR_freedomHouselag1_sq |   .0501715   .0601932     0.83   0.405    -.0678049     .168148
                      gdp_pc_constantUS2010lag1 |   .5651616   .4982902     1.13   0.257    -.4114692    1.541792
                   gdp_pc_constantUS2010lag1_sq |   -.186423    .110749    -1.68   0.092     -.403487    .0306411
                                      KOFGIlag1 |   .0160276   .0676808     0.24   0.813    -.1166243    .1486796
                                   KOFGIlag1_sq |   .0004613   .0006857     0.67   0.501    -.0008826    .0018053
                                 populationlag1 |   .5775184   .2496061     2.31   0.021     .0882994    1.066737
                                          _cons |  -5.696616   1.735481    -3.28   0.001    -9.098097   -2.295135
------------------------------------------------+----------------------------------------------------------------
                                       /lnalpha |    2.38286   .1459672                       2.09677    2.668951
------------------------------------------------+----------------------------------------------------------------
                                          alpha |   10.83585   1.581679                      8.139836    14.42483
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,210 -640.7467  -606.3006      16    1244.601   1326.175
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS6.doc", append ///
>  ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS6.doc
dir : seeout

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0(1)10) ) post

Predictive margins                              Number of obs     =      1,210
Model VCE    : Robust

Expression   : Predicted number of events, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           1

3._at        : RESTRI~mlag1    =           2

4._at        : RESTRI~mlag1    =           3

5._at        : RESTRI~mlag1    =           4

6._at        : RESTRI~mlag1    =           5

7._at        : RESTRI~mlag1    =           6

8._at        : RESTRI~mlag1    =           7

9._at        : RESTRI~mlag1    =           8

10._at       : RESTRI~mlag1    =           9

11._at       : RESTRI~mlag1    =          10

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5662929   .4906836     1.15   0.248    -.3954294    1.528015
          2  |   .7140904   .6212183     1.15   0.250     -.503475    1.931656
          3  |   .8378963   .7405379     1.13   0.258    -.6135313    2.289324
          4  |    .914855   .8200059     1.12   0.265     -.692327    2.522037
          5  |   .9294781   .8367874     1.11   0.267    -.7105951    2.569551
          6  |    .878721   .7842079     1.12   0.262    -.6582983     2.41574
          7  |   .7730147   .6749964     1.15   0.252    -.5499539    2.095983
          8  |   .6327752   .5357971     1.18   0.238    -.4173677    1.682918
          9  |   .4819879   .3961446     1.22   0.224    -.2944412    1.258417
         10  |   .3416234   .2777204     1.23   0.219    -.2026986    .8859454
         11  |   .2253119   .1889135     1.19   0.233    -.1449517    .5955755
------------------------------------------------------------------------------

. marginsplot, recast(line) recastci(rarea)  ///
> plotopt(color(gs0) lwidth(1) ) ///
> ciopt(color(gs6) fintensity(10) lcolor(gs16) ) ///
> xtitle("Count of restriction types", size(large)) ///
> ytitle("Predicted number of media shaming events", size(large)) ///
> title("Model 2 (INGOs)", size(large)) ///
> scheme(s1mono)

  Variables that uniquely identify margins: RESTRICT_COUNTdomlag1

. graph export ".\Figures\Appendix_FigureS6_2.png", replace
(note: file .\Figures\Appendix_FigureS6_2.png not found)
(file .\Figures\Appendix_FigureS6_2.png written in PNG format)

. 
. 
. * Zero inflated negative binomial 
. set seed 2

. #delimit ;
delimiter now ;
. zinb shamingINGO c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1, 
> inflate(fhbest fhworst UCDP_armedConflictlag1 shamingINGOlag1) 
> vce(cluster cowcode);

Fitting constant-only model:

Iteration 0:   log pseudolikelihood = -782.62635  (not concave)
Iteration 1:   log pseudolikelihood = -658.31164  
Iteration 2:   log pseudolikelihood = -615.21644  (not concave)
Iteration 3:   log pseudolikelihood = -606.25176  
Iteration 4:   log pseudolikelihood = -596.06761  
Iteration 5:   log pseudolikelihood = -590.95704  
Iteration 6:   log pseudolikelihood = -590.48231  
Iteration 7:   log pseudolikelihood = -590.44353  
Iteration 8:   log pseudolikelihood = -590.44327  
Iteration 9:   log pseudolikelihood = -590.44327  

Fitting full model:

Iteration 0:   log pseudolikelihood = -590.44327  
Iteration 1:   log pseudolikelihood = -575.51979  
Iteration 2:   log pseudolikelihood = -572.88025  
Iteration 3:   log pseudolikelihood = -572.77003  
Iteration 4:   log pseudolikelihood = -572.76991  
Iteration 5:   log pseudolikelihood = -572.76991  

Zero-inflated negative binomial regression      Number of obs     =      1,182
                                                Nonzero obs       =        123
                                                Zero obs          =      1,059

Inflation model      = logit                    Wald chi2(14)     =      43.13
Log pseudolikelihood = -572.7699                Prob > chi2       =     0.0001

                                                                 (Std. Err. adjusted for 147 clusters in cowcode)
-----------------------------------------------------------------------------------------------------------------
                                                |               Robust
                                    shamingINGO |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
shamingINGO                                     |
                          RESTRICT_COUNTdomlag1 |   .2061893   .1330374     1.55   0.121    -.0545593    .4669379
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0290637   .0147394    -1.97   0.049    -.0579524   -.0001749
                                                |
                                      PTS_Slag1 |   .3895593   .1503685     2.59   0.010     .0948424    .6842762
                                   hrgroupslag1 |  -.0014808   .0047043    -0.31   0.753     -.010701    .0077393
                                     hrnewslag1 |   .1972925   .0657657     3.00   0.003      .068394     .326191
                     protest_ClarkRegan_loglag1 |   .1538067   .1427021     1.08   0.281    -.1258843    .4334978
                         UCDP_armedConflictlag1 |  -.4009597   .4896285    -0.82   0.413    -1.360614    .5586945
                            PR_freedomHouselag1 |  -.0057073   .6353206    -0.01   0.993    -1.250913    1.239498
                         PR_freedomHouselag1_sq |    .029351   .0749777     0.39   0.695    -.1176026    .1763046
                      gdp_pc_constantUS2010lag1 |   .6037872   .4796785     1.26   0.208    -.3363653     1.54394
                   gdp_pc_constantUS2010lag1_sq |  -.1722872    .096133    -1.79   0.073    -.3607043      .01613
                                      KOFGIlag1 |  -.0212026   .0623814    -0.34   0.734     -.143468    .1010627
                                   KOFGIlag1_sq |    .000521   .0005779     0.90   0.367    -.0006116    .0016536
                                 populationlag1 |   .2949078   .2450004     1.20   0.229    -.1852841    .7750998
                                          _cons |  -2.200548   1.906271    -1.15   0.248    -5.936771    1.535675
------------------------------------------------+----------------------------------------------------------------
inflate                                         |
                                         fhbest |   .0846782   .4664263     0.18   0.856    -.8295005    .9988569
                                        fhworst |   .0728148   .4712953     0.15   0.877     -.850907    .9965365
                         UCDP_armedConflictlag1 |  -.6152689   .4396233    -1.40   0.162    -1.476915    .2463769
                                shamingINGOlag1 |  -.9173746   .2967214    -3.09   0.002    -1.498938   -.3358113
                                          _cons |   1.465622   .3509116     4.18   0.000     .7778476    2.153396
------------------------------------------------+----------------------------------------------------------------
                                       /lnalpha |   .6525785   .3584129     1.82   0.069    -.0498978    1.355055
------------------------------------------------+----------------------------------------------------------------
                                          alpha |   1.920486   .6883271                      .9513266    3.876974
-----------------------------------------------------------------------------------------------------------------

.  // vuong ;
> #delimit cr
delimiter now cr
. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |      1,182 -590.4433  -572.7699      21     1187.54   1294.114
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS6.doc", append ///
>  ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS6.doc
dir : seeout

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0 4)) contrast(atcontrast(r)) //   .1116074   .1444988     -.1716051      .39482

Contrasts of predictive margins
Model VCE    : Robust

Expression   : Predicted number of events, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           4

------------------------------------------------
             |         df        chi2     P>chi2
-------------+----------------------------------
         _at |          1        1.07     0.3010
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
         _at |
   (2 vs 1)  |   .1325317   .1281342     -.1186066    .3836701
--------------------------------------------------------------

. margins, at(RESTRICT_COUNTdomlag1 = (4 10)) contrast(atcontrast(r)) //   -.3770978   .1572203      -.685244   -.0689516

Contrasts of predictive margins
Model VCE    : Robust

Expression   : Predicted number of events, predict()

1._at        : RESTRI~mlag1    =           4

2._at        : RESTRI~mlag1    =          10

------------------------------------------------
             |         df        chi2     P>chi2
-------------+----------------------------------
         _at |          1        3.97     0.0464
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
         _at |
   (2 vs 1)  |  -.3070805   .1541477     -.6092044   -.0049565
--------------------------------------------------------------

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0(1)10) ) post

Predictive margins                              Number of obs     =      1,182
Model VCE    : Robust

Expression   : Predicted number of events, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           1

3._at        : RESTRI~mlag1    =           2

4._at        : RESTRI~mlag1    =           3

5._at        : RESTRI~mlag1    =           4

6._at        : RESTRI~mlag1    =           5

7._at        : RESTRI~mlag1    =           6

8._at        : RESTRI~mlag1    =           7

9._at        : RESTRI~mlag1    =           8

10._at       : RESTRI~mlag1    =           9

11._at       : RESTRI~mlag1    =          10

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .3061098   .0791928     3.87   0.000     .1508947    .4613249
          2  |   .3654281   .0866337     4.22   0.000     .1956292     .535227
          3  |   .4116065   .1031451     3.99   0.000     .2094459    .6137671
          4  |   .4374397   .1181502     3.70   0.000     .2058695    .6690098
          5  |   .4386415   .1226083     3.58   0.000     .1983337    .6789493
          6  |   .4150085   .1136609     3.65   0.000     .1922372    .6377797
          7  |   .3704757   .0947962     3.91   0.000     .1846787    .5562728
          8  |   .3120457   .0742561     4.20   0.000     .1665065     .457585
          9  |   .2479889   .0611424     4.06   0.000      .128152    .3678259
         10  |   .1859525   .0574609     3.24   0.001     .0733313    .2985737
         11  |   .1315611   .0563514     2.33   0.020     .0211144    .2420077
------------------------------------------------------------------------------

. marginsplot, recast(line) recastci(rarea)  ///
> plotopt(color(gs0) lwidth(1) ) ///
> ciopt(color(gs6) fintensity(10) lcolor(gs16) ) ///
> xtitle("Count of restriction types", size(large)) ///
> ytitle("Predicted number of media shaming events", size(large)) ///
> title("Model 3 (INGOs)", size(large)) ///
> scheme(s1mono)

  Variables that uniquely identify margins: RESTRICT_COUNTdomlag1

. graph export ".\Figures\Appendix_Figure6_3.png", replace
(note: file .\Figures\Appendix_Figure6_3.png not found)
(file .\Figures\Appendix_Figure6_3.png written in PNG format)

. 
. 
. 
. 
. 
. ***************************************************************************
. **** Appendix S7: Fixed effects models
. ***************************************************************************
. 
. xtset cowcode YEAR
       panel variable:  cowcode (unbalanced)
        time variable:  YEAR, 1986 to 2016, but with a gap
                delta:  1 unit

. 
. * Poisson RANDOM EFFECTS
. #delimit ;
delimiter now ;
. xtpoisson urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1, re;

Fitting Poisson model:

Iteration 0:   log likelihood = -5146.6988  
Iteration 1:   log likelihood = -5146.3488  
Iteration 2:   log likelihood = -5146.3487  

Fitting full model:

Iteration 0:   log likelihood =  -2866.058  
Iteration 1:   log likelihood = -2622.0975  
Iteration 2:   log likelihood = -2613.8694  
Iteration 3:   log likelihood = -2613.8402  
Iteration 4:   log likelihood = -2613.8402  

Random-effects Poisson regression               Number of obs     =      1,691
Group variable: cowcode                         Number of groups  =        171

Random effects u_i ~ Gamma                      Obs per group:
                                                              min =          1
                                                              avg =        9.9
                                                              max =         10

                                                Wald chi2(2)      =      82.42
Log likelihood  = -2613.8402                    Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------------------------------
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |   .2350741   .0288588     8.15   0.000     .1785119    .2916363
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0201728   .0031338    -6.44   0.000    -.0263149   -.0140306
                                                |
                                          _cons |   .3259041   .1344243     2.42   0.015     .0624374    .5893708
------------------------------------------------+----------------------------------------------------------------
                                       /lnalpha |   1.013707   .1092297                      .7996208    1.227793
------------------------------------------------+----------------------------------------------------------------
                                          alpha |   2.755798   .3010149                      2.224697    3.413688
-----------------------------------------------------------------------------------------------------------------
LR test of alpha=0: chibar2(01) = 5065.02              Prob >= chibar2 = 0.000

. #delimit cr
delimiter now cr
. estimates store random

. * Poisson FIXED EFFECTS
. #delimit ;
delimiter now ;
. xtpoisson urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1, fe;
note: 1 group (1 obs) dropped because of only one obs per group
note: 38 groups (380 obs) dropped because of all zero outcomes

Iteration 0:   log likelihood = -2042.7541  
Iteration 1:   log likelihood = -2009.7723  
Iteration 2:   log likelihood = -2009.7352  
Iteration 3:   log likelihood = -2009.7352  

Conditional fixed-effects Poisson regression    Number of obs     =      1,310
Group variable: cowcode                         Number of groups  =        132

                                                Obs per group:
                                                              min =          7
                                                              avg =        9.9
                                                              max =         10

                                                Wald chi2(2)      =      64.33
Log likelihood  = -2009.7352                    Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------------------------------
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |   .2055632    .028955     7.10   0.000     .1488124     .262314
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0175548   .0031688    -5.54   0.000    -.0237655    -.011344
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. estimates store fixed

. hausman random fixed

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |     random       fixed        Difference          S.E.
-------------+----------------------------------------------------------------
RESTRI~mlag1 |    .2350741     .2055632        .0295109               .
          c. |
RESTRI~mlag1#|
          c. |
RESTRI~mlag1 |   -.0201728    -.0175548        -.002618               .
------------------------------------------------------------------------------
                       b = consistent under Ho and Ha; obtained from xtpoisson
        B = inconsistent under Ha, efficient under Ho; obtained from xtpoisson

    Test:  Ho:  difference in coefficients not systematic

                  chi2(2) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =       83.15
                Prob>chi2 =      0.0000
                (V_b-V_B is not positive definite)

. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
       fixed |      1,310         .  -2009.735       2     4023.47   4033.826
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. mat list es_ic

es_ic[1,6]
                N         ll0          ll          df         AIC         BIC
fixed        1310           .  -2009.7352           2   4023.4705    4033.826

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS7.doc", replace ///
>  ctitle("Model 1") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS7.doc
dir : seeout

. 
. 
. 
. * Poisson FIXED EFFECTS
. #delimit ;
delimiter now ;
. quietly xtpoisson urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1 i.YEAR, fe;

. estimates store fixed ;

. #delimit cr
delimiter now cr
. * Poisson RANDOM EFFECTS
. #delimit ;
delimiter now ;
. quietly xtpoisson urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1 i.YEAR, re;

. estimates store random ;

. #delimit cr
delimiter now cr
. hausman fixed random

Note: the rank of the differenced variance matrix (20) does not equal the number of coefficients being tested (22); be sure this
        is what you expect, or there may be problems computing the test.  Examine the output of your estimators for anything
        unexpected and possibly consider scaling your variables so that the coefficients are on a similar scale.

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |     fixed        random       Difference          S.E.
-------------+----------------------------------------------------------------
RESTRI~mlag1 |    .1249513     .1718687       -.0469174        .0095762
          c. |
RESTRI~mlag1#|
          c. |
RESTRI~mlag1 |   -.0083303    -.0127312        .0044008        .0011246
   PTS_Slag1 |   -.0146341      .078017       -.0926511        .0114393
hrgroupslag1 |   -.0115144     -.006291       -.0052234        .0021014
  hrnewslag1 |    .0491065     .0465015        .0026049        .0029155
protes~glag1 |   -.0270892    -.0048801       -.0222092         .006709
UCDP_armed~1 |    .6408348     .6473237       -.0064888        .0241644
PR_freedom~1 |    .3568934     .4214159       -.0645226        .0406975
PR_free~1_sq |   -.0201905    -.0298201        .0096296        .0052916
gdp_pc_con~1 |    2.891392    -.1275409        3.018933        .9377616
gdp_pc_~1_sq |   -.3937534    -.0960474        -.297706        .2577346
   KOFGIlag1 |    .1538595     .1493571        .0045024        .0309334
KOFGIlag1_sq |   -.0009257    -.0011104        .0001846        .0002916
population~1 |   -.4073919    -.0663217       -.3410702         .223597
        YEAR |
       2000  |    .1612066     .1713842       -.0101776        .0134734
       2001  |   -.2459251    -.1517039       -.0942212        .0179984
       2002  |   -.0897359     .0296165       -.1193524        .0215433
       2003  |    .0079083     .0628074       -.0548991        .0495316
       2004  |   -.4540769    -.3117712       -.1423057        .0634979
       2005  |   -.6065068    -.4052552       -.2012516        .0796605
       2006  |   -.2827659    -.1864971       -.0962687        .1530617
       2007  |    -.503384    -.3241323       -.1792518        .1706259
------------------------------------------------------------------------------
                       b = consistent under Ho and Ha; obtained from xtpoisson
        B = inconsistent under Ha, efficient under Ho; obtained from xtpoisson

    Test:  Ho:  difference in coefficients not systematic

                 chi2(20) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =       91.86
                Prob>chi2 =      0.0000
                (V_b-V_B is not positive definite)

. 
. 
. ** Poisson with FE
. set seed 2

. #delimit ;
delimiter now ;
. xtpoisson urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1, fe ;
note: 33 groups (279 obs) dropped because of all zero outcomes

Iteration 0:   log likelihood = -1512.4028  
Iteration 1:   log likelihood = -1381.9782  
Iteration 2:   log likelihood =  -1380.195  
Iteration 3:   log likelihood = -1380.1948  
Iteration 4:   log likelihood = -1380.1948  

Conditional fixed-effects Poisson regression    Number of obs     =        971
Group variable: cowcode                         Number of groups  =        114

                                                Obs per group:
                                                              min =          5
                                                              avg =        8.5
                                                              max =          9

                                                Wald chi2(14)     =     243.76
Log likelihood  = -1380.1948                    Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------------------------------
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |   .1426613   .0383155     3.72   0.000     .0675643    .2177583
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0097299   .0041023    -2.37   0.018    -.0177702   -.0016896
                                                |
                                      PTS_Slag1 |  -.0245872   .0534684    -0.46   0.646    -.1293832    .0802089
                                   hrgroupslag1 |  -.0105733   .0014828    -7.13   0.000    -.0134795   -.0076671
                                     hrnewslag1 |   .0471817   .0102954     4.58   0.000     .0270031    .0673603
                     protest_ClarkRegan_loglag1 |  -.0393847   .0350173    -1.12   0.261    -.1080173     .029248
                         UCDP_armedConflictlag1 |   .5980214   .0921003     6.49   0.000     .4175082    .7785346
                            PR_freedomHouselag1 |    .367947   .1441638     2.55   0.011     .0853911    .6505028
                         PR_freedomHouselag1_sq |  -.0204222   .0166659    -1.23   0.220    -.0530867    .0122424
                      gdp_pc_constantUS2010lag1 |   2.031773   .9465742     2.15   0.032     .1765212    3.887024
                   gdp_pc_constantUS2010lag1_sq |  -.3338466   .2695161    -1.24   0.215    -.8620884    .1943952
                                      KOFGIlag1 |   .1218448   .0451699     2.70   0.007     .0333135    .2103762
                                   KOFGIlag1_sq |  -.0009255   .0004396    -2.11   0.035    -.0017871   -.0000639
                                 populationlag1 |   .1692623   .2397989     0.71   0.480     -.300735    .6392596
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. 
. estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |        Obs  ll(null)  ll(model)      df         AIC        BIC
-------------+---------------------------------------------------------------
           . |        971         .  -1380.195      14     2788.39   2856.686
-----------------------------------------------------------------------------
               Note: N=Obs used in calculating BIC; see [R] BIC note.

. mat es_ic = r(S)

. local AIC: display %4.1f es_ic[1,5]

. local BIC: display %4.1f es_ic[1,6]

. local LL: display %4.1f es_ic[1,3]

. outreg2 using ".\Tables\Appendix_TableS7.doc", append ///
>  ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) ///
> addtext(Log-Likelihood, `LL', BIC, `BIC', AIC, `AIC')
.\Tables\Appendix_TableS7.doc
dir : seeout

. 
. 
. 
. 
. ***************************************************************************
. **** Appendix S8: Population Average Models
. ***************************************************************************
. 
. * Negative binomial Population averaged
. #delimit ;
delimiter now ;
. xtnbreg urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1, pa corr(ar 1);

note:  observations not equally spaced
       modal spacing is delta YEAR = 1 unit
       2 groups omitted from estimation
note:  some groups have fewer than 2 observations
       not possible to estimate correlations for those groups
       1 groups omitted from estimation


Iteration 1: tolerance = .36599999
Iteration 2: tolerance = .19247065
Iteration 3: tolerance = .04814196
Iteration 4: tolerance = .002011
Iteration 5: tolerance = .00019881
Iteration 6: tolerance = 6.187e-06
Iteration 7: tolerance = 1.148e-07

GEE population-averaged model                   Number of obs     =      1,674
Group and time vars:          cowcode YEAR      Number of groups  =        168
Link:                                  log      Obs per group:
Family:             negative binomial(k=1)                    min =          7
Correlation:                         AR(1)                    avg =       10.0
                                                              max =         10
                                                Wald chi2(2)      =      53.55
Scale parameter:                         1      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------------------------------
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |   .2291575   .0346728     6.61   0.000        .1612    .2971151
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0191727   .0036851    -5.20   0.000    -.0263954   -.0119501
                                                |
                                          _cons |   .3619977   .0749589     4.83   0.000      .215081    .5089143
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. 
. outreg2 using ".\Tables\Appendix_TableS8.doc", append ///
>  ctitle("Model 3") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +) 
.\Tables\Appendix_TableS8.doc
dir : seeout

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0 6)) contrast(atcontrast(r)) //  10.06674   2.769525      4.638568    15.49491

Contrasts of adjusted predictions
Model VCE    : Conventional

Expression   : Exponentiated linear prediction, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           6

------------------------------------------------
             |         df        chi2     P>chi2
-------------+----------------------------------
         _at |          1       34.01     0.0000
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
         _at |
   (2 vs 1)  |    1.41211    .242147      .9375105    1.886709
--------------------------------------------------------------

. margins, at(RESTRICT_COUNTdomlag1 = (6 10)) contrast(atcontrast(r)) // -8.209127   2.678666     -13.45921   -2.959038

Contrasts of adjusted predictions
Model VCE    : Conventional

Expression   : Exponentiated linear prediction, predict()

1._at        : RESTRI~mlag1    =           6

2._at        : RESTRI~mlag1    =          10

------------------------------------------------
             |         df        chi2     P>chi2
-------------+----------------------------------
         _at |          1        8.33     0.0039
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
         _at |
   (2 vs 1)  |  -.7601128   .2632974     -1.276166   -.2440593
--------------------------------------------------------------

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0(1)10) )

Adjusted predictions                            Number of obs     =      1,674
Model VCE    : Conventional

Expression   : Exponentiated linear prediction, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           1

3._at        : RESTRI~mlag1    =           2

4._at        : RESTRI~mlag1    =           3

5._at        : RESTRI~mlag1    =           4

6._at        : RESTRI~mlag1    =           5

7._at        : RESTRI~mlag1    =           6

8._at        : RESTRI~mlag1    =           7

9._at        : RESTRI~mlag1    =           8

10._at       : RESTRI~mlag1    =           9

11._at       : RESTRI~mlag1    =          10

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   1.436196   .1076556    13.34   0.000     1.225195    1.647197
          2  |   1.771776   .1228312    14.42   0.000     1.531031    2.012521
          3  |    2.10354   .1554657    13.53   0.000     1.798833    2.408248
          4  |   2.403475    .195678    12.28   0.000     2.019953    2.786997
          5  |   2.642866   .2314879    11.42   0.000     2.189158    3.096574
          6  |   2.796775   .2545883    10.99   0.000     2.297791    3.295759
          7  |   2.848305    .262456    10.85   0.000     2.333901     3.36271
          8  |   2.791659   .2593388    10.76   0.000     2.283365    3.299954
          9  |   2.633207    .254965    10.33   0.000     2.133485     3.13293
         10  |   2.390311   .2588156     9.24   0.000     1.883042     2.89758
         11  |   2.088193   .2722532     7.67   0.000     1.554586    2.621799
------------------------------------------------------------------------------

. marginsplot, recast(line) recastci(rarea)  ///
> plotopt(color(gs0) lwidth(1) ) ///
> ciopt(color(gs6) fintensity(10) lcolor(gs16) ) ///
> xtitle("Count of restriction types", size(large)) ///
> ytitle("Predicted number of UAs", size(large)) ///
> title("Model 1", size(large)) ///
> scheme(s1mono)

  Variables that uniquely identify margins: RESTRICT_COUNTdomlag1

. 
. 
. * Negative binomial Population averaged
. set seed 2

. #delimit ;
delimiter now ;
. xtnbreg urgentActions c.RESTRICT_COUNTdomlag1##c.RESTRICT_COUNTdomlag1
> PTS_Slag1 hrgroupslag1 hrnewslag1 
> protest_ClarkRegan_loglag1 UCDP_armedConflictlag1 
> PR_freedomHouselag1 PR_freedomHouselag1_sq 
> gdp_pc_constantUS2010lag1 gdp_pc_constantUS2010lag1_sq 
> KOFGIlag1 KOFGIlag1_sq populationlag1, pa corr(ar 1) tol(0.01) ;

Iteration 1: tolerance = .14503651
Iteration 2: tolerance = .06137907
Iteration 3: tolerance = .03201912
Iteration 4: tolerance = .01395214
Iteration 5: tolerance = .00860685

GEE population-averaged model                   Number of obs     =      1,250
Group and time vars:          cowcode YEAR      Number of groups  =        147
Link:                                  log      Obs per group:
Family:             negative binomial(k=1)                    min =          5
Correlation:                         AR(1)                    avg =        8.5
                                                              max =          9
                                                Wald chi2(14)     =     240.09
Scale parameter:                         1      Prob > chi2       =     0.0000

-----------------------------------------------------------------------------------------------------------------
                                  urgentActions |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------------------------------+----------------------------------------------------------------
                          RESTRICT_COUNTdomlag1 |   .3003233   .0526514     5.70   0.000     .1971284    .4035183
                                                |
c.RESTRICT_COUNTdomlag1#c.RESTRICT_COUNTdomlag1 |  -.0219489   .0055385    -3.96   0.000    -.0328042   -.0110937
                                                |
                                      PTS_Slag1 |   .2403442   .0618903     3.88   0.000     .1190415    .3616469
                                   hrgroupslag1 |   .0033443    .001804     1.85   0.064    -.0001914    .0068801
                                     hrnewslag1 |   .0347969   .0295596     1.18   0.239    -.0231389    .0927327
                     protest_ClarkRegan_loglag1 |   .1156716   .0455231     2.54   0.011      .026448    .2048952
                         UCDP_armedConflictlag1 |   .5915911   .1251069     4.73   0.000     .3463861    .8367962
                            PR_freedomHouselag1 |   .5230296   .1711134     3.06   0.002     .1876534    .8584057
                         PR_freedomHouselag1_sq |  -.0471971   .0195376    -2.42   0.016    -.0854902    -.008904
                      gdp_pc_constantUS2010lag1 |   .6173458   .2605772     2.37   0.018     .1066239    1.128068
                   gdp_pc_constantUS2010lag1_sq |  -.3251569   .1174678    -2.77   0.006    -.5553896   -.0949241
                                      KOFGIlag1 |   .1557797   .0362159     4.30   0.000     .0847978    .2267617
                                   KOFGIlag1_sq |  -.0014682   .0003533    -4.16   0.000    -.0021607   -.0007757
                                 populationlag1 |   .0518341   .1153094     0.45   0.653    -.1741681    .2778363
                                          _cons |  -5.833266   .9788107    -5.96   0.000      -7.7517   -3.914833
-----------------------------------------------------------------------------------------------------------------

. #delimit cr
delimiter now cr
. 
. outreg2 using ".\Tables\Appendix_TableS8.doc", append ///
>  ctitle("Model 2") label  eqdrop(lnalpha) dec(3) ///
> alpha(0.001, 0.01, 0.05, 0.1) symbol(***,**, *, +)
.\Tables\Appendix_TableS8.doc
dir : seeout

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0 8)) contrast(atcontrast(r)) // .324735    .172486     -.0133314    .6628013

Contrasts of predictive margins
Model VCE    : Conventional

Expression   : Exponentiated linear prediction, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           8

------------------------------------------------
             |         df        chi2     P>chi2
-------------+----------------------------------
         _at |          1       24.16     0.0000
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
         _at |
   (2 vs 1)  |   1.850954   .3765525      1.112924    2.588983
--------------------------------------------------------------

. margins, at(RESTRICT_COUNTdomlag1 = (8 10)) contrast(atcontrast(r)) //-.0949212   .1362447      -.361956    .1721136

Contrasts of predictive margins
Model VCE    : Conventional

Expression   : Exponentiated linear prediction, predict()

1._at        : RESTRI~mlag1    =           8

2._at        : RESTRI~mlag1    =          10

------------------------------------------------
             |         df        chi2     P>chi2
-------------+----------------------------------
         _at |          1        4.20     0.0404
------------------------------------------------

--------------------------------------------------------------
             |            Delta-method
             |   Contrast   Std. Err.     [95% Conf. Interval]
-------------+------------------------------------------------
         _at |
   (2 vs 1)  |  -.5061508   .2470075     -.9902766    -.022025
--------------------------------------------------------------

. 
. margins, at(RESTRICT_COUNTdomlag1 = (0(1)10) )

Predictive margins                              Number of obs     =      1,250
Model VCE    : Conventional

Expression   : Exponentiated linear prediction, predict()

1._at        : RESTRI~mlag1    =           0

2._at        : RESTRI~mlag1    =           1

3._at        : RESTRI~mlag1    =           2

4._at        : RESTRI~mlag1    =           3

5._at        : RESTRI~mlag1    =           4

6._at        : RESTRI~mlag1    =           5

7._at        : RESTRI~mlag1    =           6

8._at        : RESTRI~mlag1    =           7

9._at        : RESTRI~mlag1    =           8

10._at       : RESTRI~mlag1    =           9

11._at       : RESTRI~mlag1    =          10

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   1.080877   .1046692    10.33   0.000     .8757289    1.286025
          2  |   1.427817   .1085873    13.15   0.000      1.21499    1.640644
          3  |   1.805113   .1389398    12.99   0.000     1.532796     2.07743
          4  |   2.184096   .1908387    11.44   0.000     1.810059    2.558133
          5  |   2.529149   .2463312    10.27   0.000     2.046348    3.011949
          6  |   2.802931   .2911091     9.63   0.000     2.232367    3.373494
          7  |   2.972938    .319921     9.29   0.000     2.345904    3.599971
          8  |   3.017829   .3393788     8.89   0.000     2.352659    3.682999
          9  |    2.93183   .3650447     8.03   0.000     2.216356    3.647305
         10  |   2.725953   .4079944     6.68   0.000     1.926299    3.525608
         11  |   2.425679   .4623469     5.25   0.000     1.519496    3.331863
------------------------------------------------------------------------------

. marginsplot, recast(line) recastci(rarea)  ///
> plotopt(color(gs0) lwidth(1) ) ///
> ciopt(color(gs6) fintensity(10) lcolor(gs16) ) ///
> xtitle("Count of restriction types", size(large)) ///
> ytitle("Predicted number of UAs", size(large)) ///
> title("Model 4", size(large)) ///
> scheme(s1mono)

  Variables that uniquely identify margins: RESTRICT_COUNTdomlag1

. 
. 
. log close
      name:  <unnamed>
       log:  C:\Users\hanna\Dropbox\PC\Documents\PaperProjects\Paper-Effective resistance\Code\ReplicationMaterial\Output_Append
> ices.log
  log type:  text
 closed on:  17 Oct 2019, 10:33:50
--------------------------------------------------------------------------------------------------------------------------------
